Optimizacion de Combustion ARPEL

132
Optimization of Combustion in Boilers and Furnaces Projects Documents Home

Transcript of Optimizacion de Combustion ARPEL

Page 1: Optimizacion de Combustion ARPEL

Optimization of Combustion inBoilers and Furnaces

Projects

Documents

Home

Page 2: Optimizacion de Combustion ARPEL

ARPEL Environmental Guide No. 29Optimization of Combustion in Boilers and FurnacesJune 30, 2000

FundingThis document has been exclusively prepared for the ARPEL Environmental Project Phase 2. The Projectwas funded by the Canadian International Development Agency (CIDA) and co-managed by theEnvironmental Services Association of Alberta (ESAA) and the Regional Association of Oil and NaturalGas Companies in Latin America and the Caribbean (ARPEL).

Environmental Services Association of Alberta#1710, 10303 Jasper AvenueEdmonton, Alberta T5J 3N6Tel: (780) 429-6363Fax: (780) 429-4249E-mail: [email protected]

Regional Association of Oil and Natural GasCompanies in Latin America and the CaribbeanJavier de Viana 2345CP 11200 Montevideo, URUGUAYTel.: (598-2) 410-6993 � 410-7454Fax: (598-2) 410-9207E-mail: [email protected]

CopyrightThe copyright in this document or product, whether in print or electronically stored on a CD or diskette orotherwise (the "Protected Work") is held by the Environmental Services Association of Alberta (ESAA).The Regional Association of Oil and Natural Gas Companies in Latin America and the Caribbean (ARPEL)has been granted a license to copy, distribute and reproduce this Protected Work on a cost-recovery andnon-commercial basis. This Protected Work shall not, in whole or in part, be copied, photocopied,reproduced, translated or reproduced to any electronic means or machine-readable form without priorconsent in writing from ESAA. Any copy of this Protected Work made under such consent must includethis copyright notice.

AuthorsThese Guidelines have been prepared upon request of ARPEL and its Environment, Health and SafetyCommittee by:Optimum Energy Management Inc.921 - 18 Avenue S.W.Calgary, Alberta T2T 0H2Tel.: (403) 215-6580Fax: (403) 215-6599E-mail: [email protected] were assisted in detailed drafting and revision by the ARPEL Energy Efficiency ProjectWorking Group.

ReviewersJaime F. George ECOPETROLJorge Velasco Urquiza CUPETWinston Charles PETROTRINManuel Olivares Paez PEMEX

Miguel Moyano ARPEL Executive Secretariat

Oscar González Environmental Services Association of Alberta

DisclaimerWhilst every effort has been made to ensure the accuracy of the information contained in this publication,neither ARPEL, nor any of its member, nor ESAA, nor any of its member companies, nor CIDA, nor theconsultants, will assume liability for any use made thereof.

Page 3: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29i

ABSTRACT

This guideline has been prepared to assist the members of ARPEL in optimizing the operationof their fired heaters, boilers and incinerators. It is not intended to be an engineering designmanual.

Chapters 1 and 2 discuss the two principal components of combustion – fuel and air. Data aresupplied so that the readers can prepare heat balances around their fired equipment. Chapter 3presents methods for estimating thermal efficiency, and where energy losses are beingexperienced. Examples are included in order to provide an indication of the impact of variousenergy management initiatives.

Heater/boiler configuration and burner construction are extremely important factors to considerwhen evaluating fired equipment. Brief descriptions of the major classifications of heaters andburners are provided in Chapter 4 to illustrate the wide range of options that are available.

Chapters 5, 6 and 7 demonstrate ways of improving the efficiency of heaters and boilers.Chapter 5 looks at the daily operation and discusses analyses, monitoring and firing practicesthat should be conducted as a part of normal operating procedures.

Heater optimizations should not be confined to reducing the fuel consumed in order to supplyheat to the process. Steps should be taken to minimize process heat requirements. With this inmind, Chapter 6 deals with ways to improve heat transfer using exchangers and options forreducing heat input into distillation columns, which ultimately leads to a reduction in heater duty.The chapter also lists energy management items concerning steam use in order to reduce firedboiler loads.

Chapter 7 discusses opportunities for reducing flue gas temperatures by using waste heatrecovery to preheat combustion air; boiler feedwater and circulating heat media. Heat sinks forcyclic operations are another option for capturing unused heat.

Flue gas emissions are a major source of air contaminants. Chapter 8 lists factors so thatARPEL members can prepare emission inventories resulting from the operation of their firedheaters and boilers. Factors for emission-control technologies are also provided.

Chapters 9, 10 and 11 describe options for controlling emissions by changing combustion(Chapter 9), by treating flue gas (Chapter 10) and by improving fuel quality (Chapter 11). Theair contaminants of concern are NOx, SOx, particulates and carbon monoxide. The impact ofswitching from oil to gas-firing, of upgrading heavy fuel oil through blending and of treating fuelare presented. The guideline closes with a discussion on the feasibility of recovering flare gasfor use as fuel.

Note any mention made to manufacturers names or trademarks of equipment and/or processesin this document does not represent an endorsement neither by the authors nor by ARPEL.

Page 4: Optimizacion de Combustion ARPEL
Page 5: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29iii

TABLE OF CONTENTS

1.0 FUEL PROPERTIES..........................................................................................................1

1.1 Heating Values ..........................................................................................................1

1.2 Contaminants ............................................................................................................7

1.3 Flame Temperature...................................................................................................8

2.0 AIR REQUIREMENTS .....................................................................................................13

2.1 Stoichiometric Air ....................................................................................................17

2.1.1 Excess Air ............................................................................................................19

2.2 Relationship Between Excess Oxygen and Excess Air............................................19

2.3 Oxygen Enrichment .................................................................................................21

3.0 HEATER EFFICIENCY ....................................................................................................23

3.1 Effect of Process Variables......................................................................................24

3.2 Energy Losses.........................................................................................................27

3.3 Boiler Efficiency.......................................................................................................30

3.4 Examples ................................................................................................................31

4.0 TYPES OF FIRED EQUIPMENT......................................................................................35

4.1 Direct-Fired Heaters ................................................................................................35

4.2 Fired Boilers ............................................................................................................37

4.3 Incinerators .............................................................................................................40

4.3.1 Incinerators for Process Wastes...........................................................................40

4.3.2 Incinerators for Medical Wastes ...........................................................................41

4.4 Other Heaters..........................................................................................................41

4.5 Burners....................................................................................................................42

4.6 Design Criteria.........................................................................................................44

5.0 MONITORING PERFORMANCE .....................................................................................45

Page 6: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 iv

5.1 Fuel Analysis ...........................................................................................................45

5.1.1 Wobbe Index ........................................................................................................45

5.2 Stack Gas Monitoring ..............................................................................................46

5.3 Firing Practices........................................................................................................47

6.0 ENERGY REDUCTION TECHNIQUES............................................................................57

6.1 Minimizing Heater Duties.........................................................................................57

6.1.1 Heat Exchange.....................................................................................................57

6.1.2 Effective Distillation ..............................................................................................58

6.2 Steam Systems .......................................................................................................64

6.3 Insulation.................................................................................................................66

7.0 ENERGY RECOVERY TECHNIQUES.............................................................................67

7.1 Modifying Radiant and Convection Sections............................................................68

7.2 Conversion to Forced-Draft Air Supply ....................................................................71

7.2.1 Forced-Draft Air Supply and Low-NOx Burners.....................................................71

7.3 Waste Heat Recovery..............................................................................................72

7.3.1 Combustion Air Preheat .......................................................................................73

7.3.2 Boiler Feedwater Preheat.....................................................................................75

7.3.3 Fuel Preheat.........................................................................................................75

7.3.4 Circulating-Liquid Heat Exchangers......................................................................75

7.3.5 Generation of Steam ............................................................................................76

7.3.6 Waste Heat Recovery from Incinerators ...............................................................76

7.4 Heat Sinks for Cyclic Operations .............................................................................76

8.0 EMISSIONS FROM HEATERS AND BOILERS...............................................................79

8.1 Emission Factors.....................................................................................................79

8.2 Effect of Heater Size on Emissions..........................................................................82

8.3 Effect of Controls on Emissions...............................................................................83

Page 7: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29v

9.0 CONTROLLING EMISSIONS – COMBUSTION CONTROLS .........................................85

9.1 Fuel Switching .........................................................................................................85

9.2 Excess Air vs. CO, NOx, Efficiency ..........................................................................86

9.3 NOx Abatement .......................................................................................................88

9.3.1 Fuel NOx...............................................................................................................89

9.3.2 Thermal NOx.........................................................................................................89

9.3.3 Flue Gas Recirculation .........................................................................................89

9.3.4 Staged-Burners and Combustion..........................................................................91

9.3.5 Low NOx Burners..................................................................................................94

9.3.6 Diluent Injection....................................................................................................96

10.0 CONTROLLING EMISSIONS-POST-COMBUSTION CONTROLS .................................99

10.1 NOx ........................................................................................................................99

10.1.1 NOx Abatement Considerations ....................................................................101

10.2 SOx ......................................................................................................................103

10.2.1 SOx and NOx.................................................................................................105

10.3 Particulates ...........................................................................................................106

11.0 FUEL IMPROVEMENT ..................................................................................................107

11.1 Sulfur Removal......................................................................................................107

11.1.1 Blending .......................................................................................................107

11.1.2 Hydrotreating ................................................................................................108

11.1.3 Fuel Gas Sweetening ...................................................................................110

11.2 Flare Gas Recovery...............................................................................................110

12.0 BIBLIOGRAPHY............................................................................................................113

Page 8: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 vi

LIST OF TABLES

TABLE 1.1: HEATS OF COMBUSTION AT 25°C OF LIGHT HYDROCARBONS, GJ/TONNE ..2

TABLE 1.2: HEATS OF COMBUSTION OF LIQUID FUELS......................................................2

TABLE 1.3 ANALYSES OF COALS .........................................................................................7

TABLE 1.4: ADIABATIC FLAME TEMPERATURES OF COMMON LIGHT FUELS,°F ........9

TABLE 1.5 CALCULATED FLAME TEMPERATURE OF COMMON LIGHT FUELS,°F.......10

TABLE 2.1: COMPOSITION OF DRY AIR...............................................................................13

TABLE 2.2: ENTHALPY OF DRY AIR .....................................................................................13

TABLE 2.3: RELATIONSHIP BETWEEN ELEVATION AND BAROMETRIC PRESSURE......16

TABLE 2.4: ENTHALPY OF WATER VAPOR..........................................................................17

TABLE 2.5: STOICHIOMETRIC AIR REQUIREMENTS FOR GASEOUS FUELS ...................18

TABLE 3.1: ENTHALPY VALUES OF FUELS .........................................................................25

TABLE 3.2: ENTHALPY OF ATOMIZING STEAM ...................................................................26

TABLE 3.3: BOILER BLOWDOWN VS. LOST BOILER EFFICIENCY .....................................30

TABLE 3.4: ENTHALPY OF FLUE GAS COMPONENTS ........................................................33

TABLE 4.1: TYPICAL BATH HEATERS ..................................................................................42

TABLE 4.2: BURNER SPECIFICATIONS................................................................................43

TABLE 5.1: RECOMMENDED EXCESS AIR LEVELS, % .......................................................48

TABLE 6.1: TYPICAL TRAY EFFICIENCIES...........................................................................62

TABLE 7.1: HEAT FLUX RATES AND MASS VELOCITIES FOR REFINERY PROCESSHEATERS ............................................................................................................68

TABLE 8.1: EMISSION FACTORS FOR FIRED HEATERS (NATURAL GAS), LBS/MMBTU..79

Page 9: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29vii

TABLE 8.2: EMISSION FACTORS FOR FIRED HEATERS (REFINERY GAS), LBS/MMBTU 80

TABLE 8.3: EMISSION FACTORS FOR AIR CONTAMINANTS FROM UNCONTROLLEDRESIDUAL OIL COMBUSTION............................................................................81

TABLE 8.4: EFFECTIVENESS OF NOX CONTROL MEASURES FOR RESIDUAL FUELCOMBUSTION.....................................................................................................83

TABLE 8.5: POST-COMBUSTION SO2 CONTROLS FOR RESIDUAL OIL COMBUSTION ....84

TABLE 8.6: REMOVAL EFFICIENCY OF TECHNOLOGIES FOR CONTROLLINGEMISSIONS OF PARTICULATES........................................................................84

TABLE 9.1: EFFECT OF FUEL SWITCHING ON HEATER EMISSIONS, LBS/HOUR ............86

TABLE 11.1: EFFECT OF FUEL OIL BLENDING ON EMMISSIONS, LBS/DAY ...................108

TABLE 11.2: EFFECT OF FUEL OIL HYDROTREATING ON EMISSIONS, LBS/DAY..........110

Page 10: Optimizacion de Combustion ARPEL
Page 11: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29ix

LIST OF FIGURES

FIGURE 1.1: BASE CASE HEATS OF COMBUSTION OF LIQUID PETROLEUM FUELS........3

FIGURE 1.2: BASE CASE SULFUR CONTENT IN LIQUID FUELS ..........................................4

FIGURE 1.3: BASE CASE CONTENT OF INERT MATERIAL IN LIQUID FUELS .....................4

FIGURE 1.4: CARBON TO HYDROGEN WEIGHT RATIO........................................................5

FIGURE 1.5: EFFECT OF EXCESS AIR AND AIR TEMPERATURE UPON ADIABATICFLAME TEMPERATURE .....................................................................................9

FIGURE 1.6: DISSOCIATION OF CARBON DIOXIDE AND WATER VAPOR.........................11

FIGURE 2.1: WATER VAPOR CONTENT IN AIR, LBS OF WATER PER LB OF DRY AIR.....14

FIGURE 2.2: WET-BULB TEMPERATURE AS A FUNCTION OF AIR TEMPERATURE ANDRELATIVE HUMIDITY .......................................................................................15

FIGURE 2.3: WATER CONTENT OF AIR-CORRECTION FACTOR FOR BAROMETRICPRESSURE .......................................................................................................16

FIGURE 2.4: EXCESS AIR AS A FUNCTION OF OXYGEN IN THE FLUE GAS.....................21

FIGURE 3.1: THERMAL EFFICIENCY OF A GAS-FIRED HEATER........................................28

FIGURE 3.2: STACK TEMPERATURE AS A FUNCTION OF FLUE GAS FLOW RATE.......28

FIGURE 3.3: RADIATION AND CONVECTIVE HEAT LOSSES ..............................................29

FIGURE 4.1: TYPICAL DIRECT FIRED PROCESS HEATERS...............................................36

FIGURE 4.2: COMMON BOILER CONFIGURATIONS............................................................39

FIGURE 4.3: PERFORMANCE OF A PACKAGED BOILER....................................................40

FIGURE 5.1: TYPICAL DRAFT PROFILE IN A DIRECT FIRED HEATER...............................50

FIGURE 5.2: EMISSIONS OF HYDROGEN AND CARBON MONOXIDE UNDERSUBSTOICHIOMETRIC CONDITIONS..............................................................50

Page 12: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 x

FIGURE 5.3: THERMAL EFFICIENCY (LHV) FOR A HEATER FIRING NATURAL GAS ........51

FIGURE 5.4: ENERGY LOSSES WHEN BURNING NATURAL GAS ......................................52

FIGURE 5.5: ENERGY LOSSES WHEN BURNING NO. 6 OIL ...............................................52

FIGURE 5.6: EFFECT OF SOOT DEPOSITS ON FUEL COMBUSTION.................................54

FIGURE 5.7: EFFECT OF SCALE DEPOSITS ON FUEL COMBUSTION...............................55

FIGURE 5.8: AIR LEAKS INTO HEATERS..............................................................................55

FIGURE 6.1: EFFECT OF PRESSURE ON REBOILER DUTY ...............................................61

FIGURE 7.1: HEAT TRANSFER IN THE RADIANT SECTION OF DIRECT FIRED HEATER .70

FIGURE 7.2: ACID GAS DEW POINT AS A FUNCTION OF FUEL SULFUR/H2S CONTENT.73

FIGURE 7.3: SCHEMATIC ARRANGEMENT OF AN AIR PREHEATER.................................75

FIGURE 9.1: CARBON MONOXIDE EMISSIONS FROM HEATERS ......................................87

FIGURE 9.2: EFFECT OF OXYGEN ON HEATER EMISSIONS .............................................88

FIGURE 9.3: NOX EMISSIONS FROM HEATERS...................................................................88

FIGURE 9.4: EFFECT OF FLUE GAS RECIRCULATION ON NOX EMISSIONS.....................90

FIGURE 9.5: STAGED BURNERS ..........................................................................................92

FIGURE 9.6: EFFECT OF LOW-NOX BURNERS ON NOX EMISSIONS (GAS FIRING)..........95

FIGURE 9.7: EFFECT OF NITROGEN CONTENT IN FUEL OIL ON NOX EMISSIONS USINGSTANDARD, HIGH INTENSITY / LOW NOX HIGH INTENSITY BURNERS.......95

FIGURE 9.8: EFFECT OF AIR PREHEAT ON NOX EMISSIONS (OIL FIRING) USINGSTANDARD HIGH INTENSITY AND LOW NOX HIGH INTENSITY BURNERS .96

FIGURE 9.9: EFFECT OF STEAM INJECTION ON NOX EMISSIONS ....................................97

FIGURE 10.1: NOX ABATEMENT TECHNOLOGIES.............................................................102

Page 13: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29xi

FIGURE 11.1: HYPOTHETICAL FLARING PATTERNS........................................................111

Page 14: Optimizacion de Combustion ARPEL
Page 15: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 291

1.0 FUEL PROPERTIES

The overwhelming majority of fired boilers and heaters, or furnaces, that are operated by the oiland gas industry use gas or heavy fuel oil. In a few cases, crude oil is burned as a fuel and instill fewer cases, coal is burned. The properties of these various fuels will influence theoperation of the fired heaters and boilers and the emissions resulting from combustion.

1.1 Heating Values

During complete combustion, hydrocarbons will be converted to carbon dioxide andwater. Heat will be released. For example, when methane is burned, the followingreaction occurs:

CH4 + 2O2 → CO2 + 2H2O + Heat

The heat of combustion represents the heat of formation of the products of the reaction.The amount will depend upon whether the water produced by the combustion is in theform of water vapor or liquid water. In the first case, the amount of heat released at 77°F(25°C) and constant pressure is 21,502 BTU/lb of methane. In the second case, theamount of heat is 23,861 BTU/lb.

The smaller value (21,502 BTU/lb) is called the low heating value (LHV) or net heatingvalue and the greater value (23,861 BTU/lb) is called the high heating value (HHV) orgross heating value. The difference between the two heating values (2,359 BTU/lb) isthe heat of vaporization of the water formed during the combustion of one pound massof methane.

It is extremely important to be consistent when using heating values. They have aprofound effect upon heater and boiler efficiency (see Chapter 3). By conventionCanada and the United States sell natural gas in terms of its high heating value. Muchof the rest of the world uses low heating value.

Table 1.1 lists high and low heating values of a number of gases typically found innatural gas and refinery fuel gas streams.

Note that the heat of combustion of propane and heavier compounds will depend uponwhether the material is in gaseous or liquid form. Also there are no high heating valuesfor the combustion of carbon and carbon monoxide because there is no hydrogenpresent to form water. Compounds sometimes found in gas streams, such as carbondioxide, nitrogen, helium and oxygen have no heating value and have been omitted fromTable 1.1. On the other hand, these compounds do help reduce the efficiency of theheater/boiler, albeit in a relatively minor way, by requiring enthalpy that otherwise wouldbe absorbed by process fluid being heated.

Liquid fuels can be classified as either distillates or residuals. Table 1.2 lists heats ofcombustion of these fuels.

Page 16: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 2

Table 1.1: HEATS OF COMBUSTION AT 25°C OF LIGHT HYDROCARBONS, GJ/TONNE1

HHV LHV HHV LHV

Methane (g) 55.5007 50.0137 n-Heptane (g) 48.4390 44.9244

Ethane (g) 51.8791 47.4876 n-Heptane (l) 48.0738 44.5592

Propane (g) 50.3486 46.3549 Ethane (g) 50.2998 47.1620

Propane (l) 49.9857 45.9943 Propylene (g) 48.9204 45.7827

n-Butane (g) 49.5275 45.7408 1-Butene (g) 48.4576 45.3198

n-Butane (l) 49.1577 45.3710 cis-2-Butene (g) 48.5436 45.1965

iso-Butane (g) 49.4089 45.6222 Trans-2-Butene (g) 48.2575 45.1174

iso-Butane (l) 49.0693 45.2826 iso-Butene (g) 48.1854 45.0476

n-Pentane (g) 49.0135 45.3547 Hydrogen (g) 141.7876 119.9551

n-Pentane (l) 48.6460 45.1942 Carbon (s) --- 32.7659

iso-Pentane (g) 48.9042 45.2430Carbon monoxide (g)

--- 10.1032

iso-Pentane (l) 48.5599 44.8988Hydrogen sulfide (g)

16.5076 15.2051

n-Hexane (g) 48.6785 45.1035

n-Hexane (l) 48.3134 44.7360

g gas l liquid s solid

Table 1.2: HEATS OF COMBUSTION OF LIQUID FUELS2

Fuel No. Type Density Heat of Combustion

GJ/m3

1 Distillate 0.82-0.85 37.91-38.74

2 Distillate 0.86-0.90 38.91-40.28

4 Residual 0.90-0.91 40.47-40.64

5 Residual 0.92-0.95 40.98-41.70

6 Residual 0.96-0.97 42.07-42.43

No. 3 oil was a distillate fuel but has been superseded by No. 2 fuel oil. Viscosity andsulfur content are important factors when using liquid fuels. Even though Nos. 4, 5 and6 are residual oils, they frequently have distillate cutterstock added in order to meet thesulfur and viscosity specifications.

1 All except H2S from Perry; page 3-142. H2S data from North American Combustion Handbook, page 8.2 Perry; page 9-6.

Page 17: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 293

The American Petroleum Institute has issued graphs for estimating the heating value ofliquid fuels. Figure 1.1 shows how heating value is a function of the density of the fuel.These values are typical. That is, they presuppose typical contents of sulfur, inertmaterial and water. These contaminants reduce the heating value of the fuel oil. Notethat the curves for densities of 0.93 and higher, are average values of cracked and virginfuel oils and crude oils. For densities less than 0.93, the data are for crude oils only.

Figures 1.2 and 1.3 show typical levels of sulfur and inert material in liquid fuels. Figure1.4 lists the carbon-to-hydrogen weight ratio. This latter graph is useful for determiningstoichiometric air requirements.

Correction for sulfur and inert material content in fuels with a density of 0.825 or less isnegligible and the curves in Figure 1.1 at those low densities represent pure petroleumliquids.1 Note that there are no curves for typical water content. The curves in Figure1.1 assume no water in the fuel.

Figure 1.1: BASE CASE HEATS OF COMBUSTION OF LIQUID PETROLEUM FUELS2

16500

17000

17500

18000

18500

19000

19500

20000

20500

0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 1.10

Density of Fuel, kg/litre

Hea

tin

g V

alu

e, B

TU

/lb

High Heating Value

Low Heating Value

1 Maxwell; page 180.2 API Technical Data Book, Figure 14.A1.1.

Page 18: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 4

Figure 1.2: BASE CASE SULFUR CONTENT IN LIQUID FUELS1

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

0.80 0.85 0.90 0.95 1.00 1.05 1.10

Density of Fuel, kg/litre

Wei

gh

t %

Su

lph

ur

Figure 1.3: BASE CASE CONTENT OF INERT MATERIAL IN LIQUID FUELS2

0.6

0.7

0.8

0.9

1.0

1.1

1.2

0.80 0.85 0.90 0.95 1.00 1.05 1.10

Density of Fuel, kg/litre

Wei

gh

t %

Iner

t M

ater

ial

1 API Technical Data Book, Table 14-0.1.2 API Technical Data Book; Table 14-0.1.

Page 19: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 295

Figure 1.4: CARBON TO HYDROGEN WEIGHT RATIO1

6.0

6.5

7.0

7.5

8.0

8.5

9.0

0.80 0.85 0.90 0.95 1.00 1.05 1.10

Density of Fuel, kg/litre

C:H

Wei

gh

t R

atio

In most cases, the liquid fuel burned at a facility does not have the characteristics listedin Figures 1.2 and 1.3. The following equations can be used to adjust the heating valuesprovided in Figure 1.1.

High Heating Value2

H = H0 – 0.01 * H0 * (% H2O + ∆ % S + ∆ % Inerts) + 40.5 * ∆ % S

where:

H = Adjusted high heating value, in BTU/lb of fuel

H0 = High heating value as provided in Figure 1.1, inBTU/lb of fuel

% H2O = Weight % water in the fuel

∆ % S = Weight % sulfur in the fuel minus the base caseweight % sulfur given in Figure 1.2

∆ % Inerts = Weight % inert material in the fuel minus the basecase weight % inert material given in Figure 1.3

1 API Technical Data Book, Table 14-0.1.2 API Technical Data Book, page 14-4.

Page 20: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 6

Low Heating Value1

L = L0 – 0.01 * L0 * (% H2O + ∆ % S + ∆ % Inerts) + 40.5 * ∆ % S – 10.53 *% H2O

where:

L = Adjusted low heating value, in BTU/lb of fuel

L0 = Low heating value as provided in Figure 1.1, inBTU/lb of fuel

% H2O = Weight % water in the fuel

∆ % S = Weight % sulfur in the fuel minus the base caseweight % sulfur given in Figure 1.2

∆ % Inerts = Weight % sulfur in the fuel minus the base caseweight % sulfur given in Figure 1.2

As an example, consider a fuel oil of density 0.95, with 1.50 weight % sulfur, 1.00 weight% inert material and 0.25 weight % water. From Figure 1.1, the high heating value is18,710 BTU/lb and the low heating value is 17,655 BTU/lb. The base case heats ofcombustion, or heating values, in Figure 1.1 assume no water in the fuel. Therefore, theactual heat of combustion must be adjusted for the entire amount of water in the fuel – inthis example, 0.25 weight %. The actual sulfur content of the fuel is 1.50 weight % andthe base case content, from Figure 1.2 is 1.20 weight %. The difference, ∆ % S, is 1.50– 1.20 = 0.30 wt %. Similarly, the actual content of inert material is 1.00 weight % andthe base case, from Figure 1.3, is 0.80 weight %.

Thus, ∆ % Inerts = 1.00 – 0.80 = 0.20 weight %

H = 18,710 – 0.01 * 18,710 * (0.25 + 0.30 + 0.20) + 40.5 * 0.30 = 18,582 BTU/lb

L = 17,655 – 0.01 * 17,655 * (0.25 + 0.30 + 0.20) + 40.5 * 0.30 – 10.53 * 0.25 = 17,532 BTU/lb

The heating value of coal is quite dependent upon the type of coal and its location. Thefollowing values in Table 1.3 are for coals mined in the USA. In the absence of site-specific data, use those in Table 1.3.

1 API Technical Data Book, page 14-4

Page 21: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 297

Table 1.3 ANALYSES OF COALS1

C H O S N Moisture Ash BTU/lb

Anthracite 86.7 2.2 2.9 0.5 0.8 2.3 6.9 13,540

Anthracite 80.9 3.3 4.2 0.5 1.0 2.1 10.1 13,480

Bituminous 84.0 4.8 5.6 0.6 1.1 3.5 3.9 14,550

76.6 5.2 6.2 1.3 1.6 2.6 9.1 13,610

79.2 5.7 10.0 0.6 1.5 3.1 3.0 14,290

68.4 5.6 16.4 1.2 1.4 8.2 7.0 12,160

71.5 5.8 14.3 2.6 1.6 7.2 4.2 12,950

62.8 5.9 17.4 4.3 1.0 12.1 8.6 11,480

63.4 5.7 18.6 2.3 1.3 12.4 8.7 11,420

Subbituminous 54.6 6.4 33.8 0.4 1.0 23.2 3.8 9,420

Lignite 42.4 6.7 43.3 0.7 1.7 34.8 6.2 7,210

In Table 1.3 the contents of moisture and ash are from the proximate analyses, whereasthe values for carbon, hydrogen, oxygen, sulfur and nitrogen are from the ultimateanalyses. All values are in weight %.

1.2 Contaminants

The impact of a number of contaminants on fuel quality has been discussed in theprevious section. Strictly speaking, carbon monoxide and hydrogen sulfide are notcontaminants because they do have heating values, although they are very lowcompared with the hydrocarbon components of the fuel.

Using the criterion from the previous paragraph, the carbon dioxide, nitrogen, helium,and inert materials in fuels are contaminants because they provide no heat. In fact theyreduce the efficiency of the fired equipment by absorbing a portion of the heat that isreleased during combustion. Wherever possible, site-specific analyses should be usedto determine more accurate estimates of the heats of combustion of the facility’s fuel(s).The appropriate ASTM test procedures (or equivalent) should be used to determine thelevels of fuel contaminants.

1 Perry; page 9-3.

Page 22: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 8

Combustion will generate acid gases, which leave the heater/boiler via the flue gas. Themost obvious of these are sulfurous and sulfuric acid and nitric acid, resulting from theburning of sulfur and nitrogen compounds in the fuel. However, there could also behydrochloric and hydrobromic acids.

It is extremely important that the facility know the analyses of its flue gas streams.These data can be used to estimate the acid gas dew point temperatures. Thesetemperatures determine the maximum amount of waste heat recovery that is possible.To reduce stack exit temperatures below the dew point will lead to condensation of theacid gas and therefore, corrosion of the stack metallurgy.

In practice, the minimum stack temperature should be set higher than the actual dewpoint in order to avoid localized corrosion. As an example, one manufacturer of wasteheat recovery equipment sets a minimum stack temperature of 177°C (350°F), eventhough the dew point of the fuel gas is in the range of 93-149°C (200-300°F).1

For a procedure for estimating acid gas dew points, the reader is referred to the article“Compute Dew Point of Acid Gases” by V. Ganapathy, Hydrocarbon Processing,February 1993, page 93. In the meantime, as a rule of thumb, gaseous fuels normallyhave acid gas dew points below 150°C (302°F). One site, burning crude oil withapproximately 6 weight % sulfur, had an acid gas dew point of 230°C (446°F).

1.3 Flame Temperature

Each fuel and its constituents will generate a flame temperature. If there were no heatloss to the surroundings and no dissociation of the products of combustion, the flametemperature would be at its maximum. All the enthalpy entering the burner in the form ofsensible heat of the combustion air, the sensible heat of the fuel and the heat ofcombustion would be transferred to the products of combustion. The result is thetheoretical flame temperature, or the adiabatic flame temperature (AFT). Table 1.4 listsreported values of adiabatic flame temperatures.

The AFT on a heater or boiler is an important parameter for the facility to know becauseit enables the facility to more accurately estimate emissions of NOx.

1 Maxim Heat Recovery Application Manual; page 7.

Page 23: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 299

Table 1.4: ADIABATIC FLAME TEMPERATURES OF COMMON LIGHT FUELS1,°F

Kunz et al

Hydrogen 4,056

Methane 3,698

Propane 3,818

Carbon Monoxide 4,311

Natural gas 3,700

H2 Plant PSA gas 3,273

The adiabatic flame temperature (AFT) is a function of the excess air rate, the gascomposition and the inlet conditions of the air and fuel. It can be calculated assumingthat the enthalpy of the products of combustion equals the sum of the enthalpy of theincoming air and fuel plus the heat of combustion. It is also assumed that kinetic andpotential energy terms are negligible and that there is no shaft work. Calculation of theAFT is a trial and error process. The heat capacity of the product streams from thecombustion temperature (frequently the heat of combustion is given at 25°C) to the AFTis a function of the final temperature, i.e., the estimated AFT.2 As an example of theeffect of excess air conditions upon the AFT, see Figure 1.5.

Figure 1.5: EFFECT OF EXCESS AIR AND AIR TEMPERATURE UPON ADIABATIC FLAMETEMPERATURE3

2800

3000

3200

3400

3600

3800

4000

4200

60 70 80 90 100 110 120 130 140 150

% of Stoichiom e tric Air

AF

T, °

F

1400°C1100°C

900°C

60°F

The adiabatic flame temperature is an important variable for two reasons. There is adirect relationship between the production of NOx and the excess oxygen and the AFT.

1 Kunz et al, Hydrocarbon Processing; November 1996; page 66.2 Smith and Van Ness; page 147.3 North American Combustion Handbook; page 10.

Page 24: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 10

Kunz et al1 generated the following formulae from their own work with steam-methanereformers (Formulae 1 and 2) and another study in the literature (Formula 3). In allcases, the x-axis is 10,000 divided by the adiabatic flame temperature in °Rankine. They-axis is the natural logarithm of the concentration of NOx in ppm divided by theconcentration of oxygen in %. In summary,

x = 10,000 / AFT, °R

y = ln (NOx, ppm / O2, %)

Low-NOx Burners Conventional General

y = 12.2-3.60x (1)

y = 12.6-3.58x (2)

y = 21.0-6.43x (3)

Secondly, high flame temperatures also cause a phenomenon called dissociation. Inthis process, the products of combustion break down in a form of reverse combustion:

Heat + CO2 → CO + O

Heat + H2O → H2 + O

The heat originally liberated in combusting carbon monoxide and hydrogen is now re-absorbed. The net result is that an equilibrium is reached, usually at temperatures in therange of 3,400-3,800°F (1,870-2,090°C). The equilibrium temperature is also known asthe calculated flame temperature. It is also a theoretical value because the effect of heatremoval and losses has not been taken into account. Calculated flame temperatures(with air at standard conditions, 100% of stoichiometric airflow and taking intoconsideration dissociation) are listed in Table 1.5. Because heaters and boilers typicallyoperate with excess air, the values in Table 1.5 are maximum values.

Table 1.5: CALCULATED FLAME TEMPERATURE OF COMMON LIGHT FUELS,°F2

Hydrogen 4,010 Propane 3,573

Carbon monoxide 3,542 n-Butane 3,583

Methane 3,484 Natural Gas 3,525

Ethane 3,540

Dissociation is relatively minor at flame temperatures less than 3000°F (1649°C).However, as indicated in Figure 1.6, the amount of dissociation, particularly for CO2

climbs rapidly. When estimating the enthalpy of the exiting flue gas, it is necessary totake into account dissociation of the products of combustion.

1 Kunz et al; pages 74, 76.2 North American Combustion Handbook; page 12.

Page 25: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2911

Figure 1.6: DISSOCIATION OF CARBON DIOXIDE AND WATER VAPOR1

0

10

20

30

40

50

60

70

3000 3500 4000 4500 5000

Temperature, °F

Am

ou

nt

of

Dis

soci

atio

n, %

H2O

CO2

Pressure at 1 atm.

1 Perry; page 9-40.

Page 26: Optimizacion de Combustion ARPEL
Page 27: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2913

2.0 AIR REQUIREMENTS

Before discussing the effect that air has upon combustion and heater/boiler efficiency it isnecessary to discuss several properties of air. The composition of air is listed in Table 2.1.

Table 2.1: COMPOSITION OF DRY AIR1

Nitrogen78.03 volume % 75.46 weight %

Oxygen 20.99 23.20

Argon 0.94 1.30

Carbon dioxide 0.03 0.04

Hydrogen 0.01 Trace

Others Trace Trace

The molecular weight is 28.964 and the density at standard temperature and pressure is 1.22kg/m3 or 0.0763 lbs/ft3. The enthalpy of dry air is given in Table 2-2. A temperature of 60°F hasbeen chosen as enthalpy = zero.

Table 2.2: ENTHALPY OF DRY AIR2

Temp.°F

EnthalpyBTU/lb

Temp.°F

EnthalpyBTU/lb

Temp.°F

EnthalpyBTU/lb

0 -14.37 180 29.16 900 210.2

20 -9.58 200 34.03 1,000 231.8

40 -4.79 250 46.21 1,100 256.2

60 0.000 300 58.43 1,200 280.5

80 4.863 400 83.05 1,300 305.4

100 9.719 500 107.9 1,400 330.2

120 14.58 600 129.5 1,500 355.5

140 19.44 700 158.4 1,600 380.7

160 24.30 800 184.1

1 North American Combustion Handbook; page 1.2 Ibid; page 344 for temperatures above 60°F and Energy Management Handbook; page 567 for

temperatures below 60°F.

Page 28: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 14

Air will contain moisture. The amount will depend upon the temperature and the relativehumidity. Figure 2.1 provides data for air at 29.92” Hg (1 atmosphere) air pressure.

At different barometric pressures the water content of air will change. To determine this effect,use the following procedure.

1. Determine the wet-bulb temperature. This can be found by knowing the air temperature(sometimes called the dry-bulb temperature) and the relative humidity. See Figure 2.2.

2. With the wet-bulb temperature and the barometric pressure, determine the moisture contentadditive correction factor for air saturated at the wet-bulb temperature when the barometricpressure differs from the standard pressure of 29.92” Hg (760 mm Hg). See Figure 2.3. Ifthe barometric pressure is not known, it can be estimated by relating it to the elevation. SeeTable 2.3. However, this method is only approximate. Barometric pressure can vary fromday to day and from season to season.

3. Calculate the difference between the dry-bulb (air) temperature and the wet-bulbtemperature, i.e., td-tw.

4. Adjust the moisture content additive correction factor (ACF) by 1% if td-tw equals 24°F. If td-tw does not equal 24°F, adjust proportionately, using the formula:

[(td-tw) / 24] * 0.01 * ACF

5. The moisture content at barometric pressure “p” is estimated using the formula

WP = W29.92” + ACF – [((td – tw) / 24) * 0.01 * ACF]

Figure 2.1: WATER VAPOR CONTENT IN AIR, lbs of water per lb of dry air1

0.00

0.01

0.02

0.03

0.04

0.05

0.06

10 20 30 40 50 60 70 80 90 100 110 120

Air Temperature, °F

lb W

ater

/lb D

ry A

ir

Parameter is Relative Humidity, %

100%

80%

60%

40%

20%

1 Perry; page 15-3.

Page 29: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2915

As an example consider the following:

The air temperature is 90°F and the relative humidity is 80%. The barometric pressure ismeasured at 27.92” Hg.

From Figure 2-1, the water content of air at a barometric pressure of 29.92” Hg is 0.025 lbswater/lb of dry air.

The dry-bulb temperature is the same as the air temperature measured by a conventionalthermometer. Using Figure 2.2, the wet-bulb temperature corresponding to an air temperatureof 90°F and a relative humidity of 80% is 85°F.

The moisture content additive correction factor for air saturated at the wet-bulb temperature isread from Figure 2.3. For a wet-bulb temperature of 85°F and a barometric pressure of 27.92”,the correction factor is 0.002 lbs water/lb of dry air. This factor must be adjusted.

The difference between the dry-bulb and wet-bulb temperatures is 90-85 = 5°F. The adjustmentis therefore, 5/24 * .01 * 0.002 = 0.0000042 lbs water/lb dry air.

The effect of reduced barometric pressure is to raise the water content of the air by 0.002 –0.0000042 = 0.001996 lbs/lb dry air. The water content is therefore, 0.025 + 0.001996 = 0.027lbs/lb dry air.

Figure 2.2: WET-BULB TEMPERATURE AS A FUNCTION OF AIR TEMPERATURE AND RELATIVEHUMIDITY1

0

20

40

60

80

100

120

10 30 50 70 90 110

Air Temperature, °F

Wet

-Bu

lb T

emp

erat

ure

, °F

Parameter is Relative Humidity100%

80%

60%

40%

20%

0%

1 Perry; pages 15-3 to 15-5.

Page 30: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 16

Table 2.3: RELATIONSHIP BETWEEN ELEVATION AND BAROMETRIC PRESSURE1

Barometric Pressure Elevation

Inches Hg Feet Metres

29.92 0 0

28.92 900 274

27.92 1,800 549

26.92 2,700 823

25.92 3,700 1,128

24.92 4,800 1,463

23.92 5,900 1,798

The enthalpy values of water vapor are listed in Table 2.4. They are adapted from SteamTables and have been adjusted to an enthalpy value of zero at 60°F.

Figure 2.3: WATER CONTENT OF AIR-CORRECTION FACTOR FOR BAROMETRIC PRESSURE2

-0.004

0

0.004

0.008

0.012

0.016

0.02

0 20 40 60 80 100 120 140

Wet-Bulb Temperature, °F

Ad

dit

ive

Co

rrec

tio

n F

acto

r,

lb W

ater

/ lb

Dry

Air

Parameter is Barometric Pressure, inches Hg

30.92"

28.92"

27.92"

26.92"

25.92"

24.92"

23.92"

29.92"

1 Perry; page 15-8.2 Perry; page 15-8.

Page 31: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2917

Table 2.4: ENTHALPY OF WATER VAPOR1

Temperature, °F Enthalpy, BTU/lbTemperature, °F

Enthalpy, BTU/lb

0 -26.1 400 151.7

20 -17.3 450 174.9

40 -8.6 500 198.2

60 0.0 600 245.8

80 8.6 700 294.2

100 17.2 800 343.3

120 26.0 900 393.2

140 34.4 1,000 443.9

160 42.7 1,100 495.5

180 50.9 1,200 547.9

200 58.6 1,300 601.8

212 63.2 1,400 655.9

250 82.0 1,500 711.6

300 104.8 1,600 766.4

350 128.2

The effect of humidity upon heater efficiency is significant. As indicated in the example,the water content in air can amount to 2.7 weight %, and in some cases even higher.The amount of heat absorbed by water vapor in the air could exceed 1-2% of the heat ofcombustion in some heaters. Another way of looking at this is to say that heating thehumidity in the air accounts for as much as 5% of the energy losses from a heater orboiler.

2.1 Stoichiometric Air

Stoichiometric air means the amount of air theoretically required to completely burn afuel. This amount is usually referred to in terms of volume of air per volume of fuel.Both volumes must be at the same temperature and pressure. The type of fuel is adeterminant factor. See Table 2.5

1 Combustion Engineering Steam Tables for temperatures up to 1200°F. For higher temperatures,

adapted from API Technical Data Book; page 14-22 and Energy Management Handbook; page 573.

Page 32: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 18

Table 2.5: STOICHIOMETRIC AIR REQUIREMENTS FOR GASEOUS FUELS1

Compound StoichiometricAir,

volume/volume

Compound StoichiometricAir,

volume/volumeHydrogen

2.39 Propylene 21.48

Carbon Monoxide 2.39 Butanes 31.02

Hydrogen Sulfide 7.16 Butylenes 28.63

Ammonia 3.58 Pentanes 38.18

Oxygen -4.78 Pentenes 35.79

Methane9.54 Hexanes 45.34

Ethane 16.70 Heptanes 52.50

Ethylene 14.32 Octanes 59.65

Propane23.86

Stoichiometric air requirements for mixtures of these gases can be calculated using thegas composition. For example, consider a fuel gas with 85% methane, 10% ethane and5% propane. The stoichiometric air requirement is:

0.85 * 9.54 + 0.10 * 16.70 + 0.05 * 23.86 = 10.97 volumes of air/volume of gas

Note that if the gas contains oxygen the amount of air required decreases. One must becareful to ensure that oxygen measured in the sample is not due to air entrapment whiletaking the sample. Gases such as nitrogen, carbon dioxide sulfur dioxide and helium donot require combustion air.

Stoichiometric air requirements are often listed in terms of volume of air/unit of mass ofliquid or solid fuel. An example is SCF air/lb of fuel. It is necessary to know the ultimateanalysis of the fuel, i.e., the amount of carbon, hydrogen, sulfur and oxygen. Theformula2 is:

SCF (60°F, 14.7 psia) = wt % C * 1.514 +wt % H * 4.54 +wt % S * 0.568 –wt % O * 0.568

This provides the stoichiometric air in SCF/lb fuel. Non-combustibles such as CO2, N2,ash and water require no air.

1 GPA Publication 2145; Figure 16-1, 1981 Revision.2 North American Combustion Handbook; page 47.

Page 33: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2919

It is important to remember that stoichiometric air assumes perfect (total) combustion. Insome cases, it is desirable to not burn a fuel to complete combustion, but in theoverwhelming majority of heaters/boilers total combustion is desired. This raises theconcept of excess air.

2.1.1 Excess Air

To ensure that there is complete combustion, a quantity of air in excess of thestoichiometric amount is added to the air/fuel mixture. This quantity is known as excessair and is expressed in % of the stoichiometric air requirement. As a rule of thumb, gas-fired heaters should have 10-15% excess air. Oil-fired heaters should have 15-25%excess air. There are two main reasons for operating heaters and boilers with excessair:

The amount of stoichiometric air is the theoretical quantity of air to achieve completecombustion. This would require perfect air/fuel mixing. However, to ensure that there isintimate contact of the air and fuel and that all the air entering the firebox comes incontact with the fuel is difficult.

Fuel quality is often changing. To operate at 0% excess air means that sometimes thereis insufficient combustion air. This results in poor energy efficiency, air quality problemsand operational upsets.

While the 10-15% excess air for gas-fired heaters is common, there are facilities thatoperate with 5-10% excess air. One company installed very sophisticated feed forwardcontrol of its fuel quality on the boilers and achieved as low as 2.5% excess air. SeeSection 5.1.1 for more detail.

In actual practice, excess air rates can climb considerably above the rule-of-thumb limitsmentioned earlier in this section. As discussed in Chapters 3 and 5, control of excess airrates can result in considerable improvement in the heater efficiency.

2.2 Relationship Between Excess Oxygen and Excess Air

Excess air is the operating parameter that must be controlled but it cannot be measureddirectly. The oxygen in the flue gas is measured and from this, the excess air isestimated. The oxygen in the flue gas is occasionally referred to as excess oxygen.This is somewhat of a misnomer. The existence of oxygen in the stack does confirmthat excess oxygen has been added but the measured quantity of oxygen is notproportional to the quantity of excess air. See Figure 2.4.

Figure 2.4 is based upon calculations for natural gas, a refinery fuel gas that hasapproximately 80% hydrogen and a sulfur-rich crude oil. The difference between thecurves is due to the fact that the flue gas analyses do not report the water content.Nevertheless, the combustion of hydrogen does consume oxygen. This changes therelationship between the reported gases – oxygen, nitrogen, carbon dioxide and sulfurdioxide – and produces a different curve.

The solid dots are data from the literature for a No. 6 Residual Fuel Oil and the solidtriangles show the relationship for a natural gas, they are taken from the literature. The

Page 34: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 20

solid squares are data for coke. This curve is essentially the same as for coal. Twoconclusions can be drawn from Figure 2.4:

1. Liquid fuels have a slightly higher excess air rate for the same concentration ofoxygen in the flue gas. Solid fuels such as coke or coal have a still higher rate.

2. Fuel streams typically consumed in oil and gas facilities are virtually identical in therelationship between excess air and oxygen in the flue gas.

It is suggested that each facility calculate an oxygen/excess air curve for its site-specificfuel. The following procedure is suggested:

♦ Determine the hydrogen to carbon ratio of the fuel. When burned in heaters andboilers, it can be assumed that the fuel is completely combusted, with the exceptionof very minor amounts of contaminants.

♦ Determine emission factors for CO, particulates, unburned hydrocarbons, SOx andNOx. They can be obtained from the ARPEL Guidelines for Atmospheric EmissionsInventory Methodologies in the Petroleum Industry.

♦ Calculate the amount of oxygen consumed in the formation of these combustionproducts.

♦ Calculate the amount of carbon in the CO and particulates (assume particulates arepure carbon). The remainder of the carbon in the feed is burned to CO2 and thehydrogen is burned to H2O.

♦ Calculate the amount of oxygen consumed in the formation of CO2 and H2O.

♦ Estimate the amount of oxygen added in the combustion/excess air. This is done byperforming a nitrogen balance around the combustion/excess air and the flue gas. (Itmay be necessary to account for any free nitrogen in the fuel.) The nitrogen can beconsidered to consist of two streams: the nitrogen associated with the combustionair and the nitrogen associated with the excess air.

♦ The amount of nitrogen associated with the excess air is determined using the ratioof nitrogen to oxygen in air and the amount of oxygen in the flue gas. The oxygen inthe flue gas is associated with the excess air (all the oxygen associated with thecombustion air is consumed in the heater/boiler).

♦ The ratio of the nitrogen associated with the excess air and the nitrogen associatedwith the combustion air gives the ratio of excess air.

Note that at low concentrations of oxygen in the flue gas the ratio of oxygen to excess airis close to the ratio observed in air – roughly 1:5. However, as the amount of oxygenincreases this ratio changes quickly. At its extreme, 21% oxygen corresponds to aninfinite excess air rate.

Page 35: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2921

Figure 2.4: EXCESS AIR AS A FUNCTION OF OXYGEN IN THE FLUE GAS1

0

50

100

150

200

250

300

350

0 2 4 6 8 10 12 14 16 18

Oxygen in Flue Gas, mol %

Exc

ess

Air

, %

Natural Gas

No. 6 Oil

Coke

Lines , from left to right, are calculated for crude oil, hydrogen-rich refinery fuel gas and natural gas

2.3 Oxygen Enrichment

The addition of combustion air- even if only the stoichiometric quantity – means that aconsiderable quantity of nitrogen (and argon, carbon dioxide and other naturalcomponents of air) must be added to the combustion process. This essentially inert gas(note, nitrogen will form thermal NOx, but for the purposes of combustion it can beconsidered as being inert) must be heated and therefore, reduces the efficiency of theprocess.

One way of reducing the amount of carrier nitrogen is to use oxygen enrichment. As arule of thumb, raising the oxygen content of the combustion air by x% will increaseheater efficiency by 0.25x % at a flue gas temperature of 600°F and by 0.5x% at a fluegas temperature of 1200°F.2 Oxygen enrichment will also improve flame velocity andwidens the flammability limits of the fuel.

Oxygen has been injected directly into the firebox using lances but better mixing controlis achieved by injecting the oxygen through the burners. Even more preferable is topremix the oxygen and the combustion air. Typically, the oxygen content is held to 25-30% in the combustion air.

1 North American Combustion Handbook; pages 54-55.2 North American Combustion Handbook, page 77.

Page 36: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 22

The use of oxygen enrichment is usually difficult to justify on economic grounds unlessthere are oxygen-generating facilities available. Moreover, there are safety factors thatmust be considered.1

1 North American Combustion Handbook, page 76.

Page 37: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2923

3.0 HEATER EFFICIENCY

Unless specifically mentioned, the term “heater” refers to fired process heaters, boilers andsteam generators. There are some differences in calculating efficiency of boilers. They arediscussed in Section 3.3.

The efficiency (%) of a heater is defined as the

Useful Heat Transferred * 100Heat Input

This equation1 is exceedingly simple to state but is often difficult to use.

First of all, there is the question of “heat input”. This is the amount of fuel multiplied by the heatof combustion. But does one use the high heating value (HHV) or the low heating value (LHV)?In the formula stated above the “useful heat transferred” is the quantity of heat transferred to theprocess fluid or steam, etc. It could also be heat transferred to the combustion air or to the fuel(see Section 3.1). This amount is constant, regardless of whether high or low heating value isused. Therefore, since the high heating value is a larger number, the efficiency (HHV basis) islower than the efficiency calculated using the low heating value.

As long as one is consistent, either heating value is appropriate. A problem may arise whencomparing actual performance with design, if the basis of calculation of the design efficiency isnot clearly stated.

The second problem with the formula for heater efficiency is that, very often, it is difficult tocalculate the amount of heat transferred to the process. This is especially true if there is partialvaporization occurring in the heater. There are manual methods, as well as computersimulation software, available to estimate the amount of vaporization but even with good fielddata there is still a reasonably large error.

One method of eliminating this potential source of error is by modifying the equation so that theuseful heat transferred is equal to the heat input less all the total losses.

Efficiency (%) = (Heat Input – Total Losses)/Heat Input * 1002

It is recommended that the efficiency be calculated using both the “useful heat transfer” and“total loss” methods. The two calculations will act as cross-checks to each other. Theavailability and quality of the field data will determine which efficiency estimate is more reliable.

It is strongly recommended that individual facilities prepare site-specific efficiency calculationsheets that list the required field and laboratory data and the calculation procedure to befollowed. The procedures range from nomographs to graphs to rigorous calculations usingcomputer simulation packages. The philosophies outlined in this document provide a basis forthese calculation sheets.

1 North American Combustion Handbook; page 56.2 Ibid, page 56.

Page 38: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 24

3.1 Effect of Process Variables

To more fully understand the effect of process variables upon heater efficiency it isnecessary to expand the “total loss” efficiency formula:

Efficiency (%) = (Heat Input – Total Losses)/Heat Input * 100

into an enthalpy balance.

This will allow one to quantify the effect of the process variables. In the present formula,the effect is buried in the term “total losses”. Note, even though the process variablesare input variables to an enthalpy balance, the term “heat input” in the formula aboverefers to the heat generated from the combustion of the fuel.

The process variables of concern are on the fuel side of the heater: heating value, airtemperature, humidity, fuel temperature, excess air and atomizing steam.

Heating Value: If the high heating value of the fuel is used, the heat lost, by notcondensing the water vapor in the flue gas, must be included in the losses. See Section3.2. If the low heating value of the fuel is used, the latent heat of the water vapor in theflue gas is omitted from the losses. The choice of which heating value to use is at thediscretion of the site, but bear in mind the following:

The site should preferably use the same basis as the design heater efficiency valueand/or the basis used in any previous comprehensive heater studies. This will allow fora more direct comparison of present operation with past results. The same rationaleapplies if the site will be comparing the performance of several heaters, i.e., in order todetermine the most efficient one.

If the stack exit temperature is approaching or even lower than the dew point of the fluegas it will be necessary to use the high heating value for the fuel.

Air Temperature: Heat from the combustion of the fuel is required to bring the air up tothe combustion temperature. Therefore, the hotter the inlet air (combustion air plusexcess air) the greater the amount of heat available for transfer to the process fluid.This results in a higher efficiency. Enthalpy values for air are provided in Table 2.2.Note that the table is for dry air only.

Humidity: The amount of water contained in air is discussed in Chapter 2. See Figures2.1, 2.2 and 2.3. The enthalpy of water vapor is listed in Table 2.4.

Fuel Temperature: As with air, heat from the combustion of the fuel is required to bringthe fuel up to the combustion temperature. Therefore, the hotter the fuel entering theheater the greater the amount of heat available for the process. The enthalpies of themore common gaseous fuels and a heavy fuel oil are provided in Table 3.1. The datafor the light hydrocarbons (gases) are from the API

Page 39: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2925

Technical Data Book. They are the values for 0 psia. Unless the actual inlet fuelpressures are very high – and this is unlikely – the errors in not using the actual pressureand in not accounting for the heat of mixing are minor and most likely within the marginof error of the field data.

The enthalpy values have been corrected to a base case enthalpy equal to zero for thegas at 60°F. This has been done purely for convenience. It is the change in enthalpythat is important.

If this convention is used, it is necessary to properly account for phase changes. Forexample, if liquid normal-butane at 60°F were heated by flue gas and vaporized, theenthalpy of the incoming liquid would have to be given an enthalpy value of –161 BTU/lb(i.e., the difference between saturated liquid and saturated vapor at 0 psia and 60°F).

For heavy fuel oil (20° API, specific gravity 0.934), enthalpy values have been providedfor 0 psia. On the basis of the procedures outlined in Chapters 4 and 7 of the APITechnical Data Book, for determining the critical properties of the fuel oil, there is noneed to apply pressure correction factors as long as the fuel oil temperature is less than800°F and the pressure is less than 300 psia. The enthalpy values have been correctedto zero for liquid at 60°F. A Watson K factor of 11.55 has been used. This is basedupon a fraction-by-fraction analysis of a crude oil with a Watson K factor of 11.8.

Table 3.1: ENTHALPY VALUES OF FUELSBTU/lb1

Temperature, °F 0 50 60 100 150 200 250 300 350 400 450 500C1 -31.0 -6.0 0.0 22.0 48.7 78.0 107.0 137.5 168.5 201.0 234.5 270.0C2 -24.0 -4.2 0.0 16.3 38.8 61.8 86.3 112.3 139.8 168.8 199.5 230.8C3 -22.0 -4.5 0.0 15.5 37.3 60.0 85.4 110.0 137.0 165.0 194.0 225.5iC4 -21.0 -3.5 0.0 16.5 37.5 61.0 85.3 111.5 138.2 167.0 197.0 228.0nC4 -23.0 -4.0 0.0 16.0 38.0 61.5 86.0 112.0 139.5 168.0 198.0 229.5iC5 -22.0 -3.8 0.0 16.0 36.5 59.5 84.0 110.0 136.5 165.0 195.0 226.0nC5 -22.2 -3.7 0.0 15.8 37.5 60.1 84.3 110.3 137.6 165.3 195.0 226.3nC6 -22.5 -4.3 0.0 15.5 36.5 59.5 83.2 108.5 135.7 164.5 193.5 224.5

C2= -20.8 -3.5 0.0 14.5 34.0 54.5 75.5 99.5 122.5 147.5 174.0 200.5

C3= -20.5 -4.0 0.0 14.5 33.5 54.5 76.0 100.0 124.5 150.0 176.0 203.5

1C4= -20.5 -3.0 0.0 14.7 34.0 55.0 77.5 101.0 125.5 152.2 178.0 207.0

cis C4= -18.7 -3.4 0.0 13.1 30.6 50.3 70.6 93.1 116.1 140.8 166.9 194.6

trans C4= -20.7 -3.7 0.0 15.0 35.0 56.3 78.7 102.3 126.3 153.1 179.3 208.3

iC4= -20.7 -4.2 0.0 15.0 35.3 56.8 79.5 104.3 129.3 154.8 182.3 210.3

H2S -14.4 -2.4 0.0 9.7 22.0 34.4 47.0 59.7 72.6 85.6 98.8 112.1H2 -203.1 -33.9 0.0 135.8 306.1 476.8 648.0 819.8 992.0 1164.7 1338.0 1511.7N2 -14.9 -2.5 0.0 9.9 22.4 34.8 47.2 59.7 72.2 84.7 97.4 110.0O2 -13.1 -2.2 0.0 8.8 19.8 30.9 42.1 53.4 64.7 76.2 87.7 99.4CO2 -11.7 -2.0 0.0 8.1 18.6 29.4 40.6 52.1 63.8 75.7 87.8 100.0CO -14.9 -2.5 0.0 9.9 22.4 34.8 47.3 59.9 72.5 85.1 97.8 110.5

20°API HFO -24.0 -4.5 0.0 17.3 40.2 64.2 89.6 117.6 146.2 175.2 206.3 238.3

Excess Air: Excess air is required in order to ensure complete combustion and stableoperation. Depending upon the fuel and the amount of firing control instrumentation,recommended excess air rates typically range from 10-30%. However, experience hasshown that rates considerably above these levels are common.

1 For hydrocarbon gases: API Technical Data Book; pages 7-52 to 7-85. For heavy fuel oil: API

Technical Data Book; pages 7-121 and 7-128. For CO, H2, CO2, N2, 02: Energy ManagementHandbook; pages 570-574.

Page 40: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 26

The presence of excess air represents a waste of energy. Heat is required to raise thisquantity of air to the combustion temperature. Enthalpy of the excess air (dry basis) islisted in Table 2.2.

Atomizing Steam: In order to improve the mixing of viscous fuels (such as heavy fueloil) with the combustion air, and to bring about vaporization of the fuel (a necessarycondition for combustion) the fuel is atomized. This can be done mechanically but acommon practice is to use air or steam. The enthalpy of atomizing air can be found inTable 2.2. The enthalpy of atomizing steam can be approximated using Table 3.2. Formore precise calculations, the reader should consult steam tables.

Note that steam tables frequently have a base enthalpy (i.e., equaling zero) for liquidwater at 32°F (0°C). At the same time, tables/graphs listing the enthalpy of water vaporhave a base of zero for the vapor at different temperatures. For instance the APITechnical Data Book uses 60°F (15.6°C); the Energy Management Handbook, byWayne C. Turner, uses 77°F (25°C). If using the steam tables, the reader has twooptions:

1. Use the steam table values, but segregate the atomizing steam in the mass balancearound the fuel side of the heater. This creates a third source of H2O in the stackgas (the other two being the humidity in the air and the product of combustion ofhydrogen). There is one problem in that steam tables may not extend to thetemperatures experienced in the flue gas or in the firebox.

2. Convert the steam table values to the same basis as the rest of the streams enteringon the fuel side. This document uses a base of zero for the gas/vapor at 60°F.Table 3.2 has been prepared on this basis.

Table 3.2: ENTHALPY OF ATOMIZING STEAM1

Corrected to same base as water vapour (Enthalpy of vapour = 0 at 60°F)

Pressure Saturated Steam Enthalpy of Superheated Steam, BTU/lbpsia Temp Enthalpy Temperature, °F

°F BTU/lb 300°F 350°F 400°F 450°F 500°F 600°F 700°F30 250.3 76.8 102.0 126.2 150.2 173.4 197.0 244.9 293.645 274.5 84.8 98.7 123.9 148.6 171.9 195.8 244.1 292.960 292.7 90.3 94.6 121.3 146.8 170.5 194.6 243.2 292.375 307.6 94.7 118.6 145.0 169.0 193.4 242.4 291.690 320.3 98.2 116.0 142.8 167.3 192.1 241.5 290.9

105 331.4 101.0 112.5 140.4 165.7 190.8 240.4 290.2120 341.3 103.4 138.2 164.2 189.5 239.6 289.5135 350.2 105.5 135.5 162.5 188.2 238.6 288.7150 358.4 107.2 132.8 160.8 186.9 237.7 287.9165 366.0 108.7 130.2 159.3 185.6 236.6 287.3180 373.1 110.0 127.4 157.5 184.3 235.6 286.5195 379.7 111.2 124.5 155.5 182.8 234.6 285.7210 385.9 112.2 121.6 153.4 181.3 233.5 284.9

It is important to remember that the enthalpy of the input streams is estimated at thepoint prior to receiving any heat input from heater flue gas. For example, if the heater isequipped with combustion air preheat (using flue gas), the input enthalpy is the inlet tothe air preheater, not the inlet to the heater.

1 Combustion Engineering Steam Tables

Page 41: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2927

3.2 Energy Losses

An energy loss is any escape/release of heat that is not recovered for use either in theheater or elsewhere. There are six main sources of loss in typical oil and gas industryheaters and boilers:

♦ Too much excess air

♦ Hot flue gas

♦ Latent heat of the water vapor

♦ Radiation and convective heat losses from the heater walls

♦ Leaks and openings in the heater construction

♦ Boiler blowdown.

Excess Air: The air entering the heater consists of combustion air (up to thestoichiometric amount) and excess air. As explained above, a certain amount of excessair is required in order to ensure total combustion. The impact of excess air on heaterefficiency is shown in Figure 3.1. It is especially significant on heaters with poor heatrecovery in the convection section.

It could be argued that the nitrogen component in the combustion air is analogous toexcess air. This is true in that heat must be supplied to the nitrogen in order to raise it tothe furnace temperature. On the other hand, the amount of excess air is a processvariable over which the operator can exercise control. Unless the heater uses oxygen-enriched air, the operator has no control over the amount of nitrogen in the combustionair.

Hot Flue Gas: In many cases, the energy leaving the heater via the flue gas is thelargest source of loss. Unlike excess air, which can be changed easily and at no cost,flue gas temperatures cannot be changed significantly without capital investment. Theimpact of flue gas temperature on heater efficiency is shown in Figure 3.1. The enthalpyvalues of flue gas components are listed in Table 3.4 at the end of this chapter.

In the previous paragraph, it was implied that flue gas temperatures can be changed tosome extent. This is shown in Figure 3.2. It is based on a study of two identical boilers,except that one is equipped with a waste heat recovery unit. These boilers are at thesame facility so that they use the same fuel and are subject to the same operatingpractices.

Note that the steam production in Figure 3.2 influences the stack temperature on bothboilers. The reason for this is presumably limiting heat flux. The amount of steam canbe considered a surrogate measurement of the amount of flue gas. The surface area(and condition) of the boiler tubes and the temperatures inside the boiler radiant andconvection sections dictate the amount of heat that can be transferred into the process(i.e., steam). When this heat transfer reaches its upper limit, the excess energy passingover the tubes leaves via the stack and results in higher stack exit temperatures.

Page 42: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 28

Figure 3.1 shows the thermal efficiency of a natural gas with a high heating value of1,000 BTU/SCF. The thermal efficiency is thus a gross (high heating value basis)efficiency. Note that the effect of heat losses from radiation and convection have notbeen included.

Figure 3.1: THERMAL EFFICIENCY OF A GAS-FIRED HEATER1

(Parameter is Percentage of Excess Air)

0

10

20

30

40

50

60

70

80

90

100

0 200 400 600 800 1000 1200 1400 1600 1800

Stack Exit Temperature, °F

Th

erm

al E

ffic

ien

cy (

HH

V),

%

Energy Datum 60°F

0%20%

40%

60%

80%

100%

150%

HHV of Gas 1000 BTU/SCF

Figure 3.2: STACK TEMPERATURE AS A FUNCTION OF FLUE GAS FLOW RATE2

250

300

350

400

450

500

550

600

40 60 80 100 120 140 160

Steam Production, 000's lb/hour

Flu

e G

as E

xit

Tem

per

atu

re, °

F

No Waste Heat Recovery

With Waste Heat Recovery

1 Gas Processors Suppliers Association (GPSA) Engineering Data Book, 9th Edition; page 8-13.2 From Author’s files.

Page 43: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2929

Latent Heat of Water Vapor: Fuels commonly used in the oil and gas industry consistof about 15-25% hydrogen by weight. Combustion will therefore, generate aconsiderable quantity of water vapor. In addition, there will be the humidity in thecombustion air/excess air and possibly atomizing steam. The thermal properties ofwater are such that the overwhelming amount of enthalpy of steam is due to the latentheat of vaporization. Since most heaters must operate at flue gas temperatures of300°F and higher, in order to avoid acid gas corrosion, there is no opportunity to capturethe latent heat of the water vapor in the flue gas. It is for this reason that the industryadopted the concept of lower heating value. However if the flue gas exit temperaturecan be in the range of 150°F, or lower, a portion of the latent heat of vaporization can becaptured and the higher heating value of the fuel (and resultant efficiency calculations)must be used.

Radiation and Convective Heat Losses: These losses are a function of the surfacetemperature of the heater relative to the surrounding air (causing radiation) and the windvelocity. See Figure 3.3.

Typically, radiation and convective heat losses amount to 2-3% of the heater firing.However, the condition and thickness of the refractory and insulation will greatly affectthis number. As shown in Figure 3.3, the surface area of the heater is a factor.

Figure 3.3: RADIATION AND CONVECTIVE HEAT LOSSES1

1

2

3

4

5

6

7

8

0 100 200 300 400 500 600

Temperature of Surface minus Temperature of Air, °F

Hea

t L

oss

Co

effi

cien

t, B

TU

/(h

r-ft

2-°F

)

0 MPH

5 MPH

10 MPH

15 MPHParameter is Wind Speed

Leaks and Openings: The amount of heat lost through leaks in the heater constructionand through openings will vary from heater to heater. Openings include the gapbetween the heater wall and incoming/outgoing heater coil piping. It also includesfirebox viewing ports left open. No quantification of these losses was found in theliterature but it is opined that the energy loss via this route is usually small. However,leaks between heater box panels and in the ductwork and access panels can besubstantial. See Section 5.3.

1 GPSA Engineering Data Book, 9th Edition; pages 8-11.

Page 44: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 30

Boiler Blowdown: Boiler efficiency calculations are discussed in Section 3.3. In termsof energy losses, there is no difference between heaters and boilers except the questionof blowdown.

Even with treating of the boiler feedwater, there will be a steady accumulation ofdissolved minerals and salts in the water. It is therefore, necessary to blow down someof the water from the boiler and replace it with fresh make-up water. The concentrationof the dissolved solids is greatest just below the water level, where the steam bubblesand water separate. A continuous blowdown arrangement is usually installed towithdraw water from this location. It is sometimes called a “top” blowdown.1

In addition, sediment and salts accumulate in the form of a sludge in the water space.This must be periodically removed using an intermittent “bottom” blowdown.2

The amount of blowdown will depend upon the incoming water quality after treatmentand the boiler pressure (the higher the steam pressure, the greater the need for cleanerboiler feedwater). Typically, blowdown rates of 5-15% are required. There will be anenergy loss if the heat in the blowdown water is not recovered for use either in the boilersystem or elsewhere.

See Table 3.3, which shows the impact of blowdown on boiler efficiency, if no heatrecovery from the blowdown is achieved.

Table 3.3: BOILER BLOWDOWN VS. LOST BOILER EFFICIENCY3

Boiler Pressure, psig

Blowdown, % 200 400 600 800

% Efficiency Lost

10 3.3 4.0 4.5 5.1

5 1.7 2.0 2.2 2.5

3.3 Boiler Efficiency

Boiler efficiency can mean “thermal efficiency”, which does not account for radiation andconvective losses, or it can mean “fuel-to-steam” efficiency, which does account forthem. Therefore, “fuel-to-steam” efficiency is a true indication of overall boilerefficiency.4 The ASME Power Test Code 4.1 specifies that the fuel-to-steam efficiencycan be calculated using either the Input-Output Method or the Heat Loss Method.

The Input-Output Method is calculated by dividing the boiler output (in energy units) bythe boiler fuel input (in energy units). The Heat Loss Method subtracts the losses via thestack and radiation/convective losses, as well as blowdown.

In summary, boiler efficiency is calculated in the same manner as heater efficiency, withthe additional energy loss term to account for blowdown.

1 Sauselein; page 55.2 Ibid; page 59.3 Garcia-Borras; page 36.4 Cleaver-Brooks, Efficiency Facts.

Page 45: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2931

3.4 Examples

The examples presented below illustrate the effects of altering the main factorsconcerning energy management of heaters and boilers. The analytical procedures anddata provided earlier in this document form the basis of this discussion. Many of thegraphs and tables can be converted to formulae using regression techniques. Thefollowing conditions exist as the base case:

Fuel Gas Fuel Oil

CO2 1.00 vol% Density 0.96

C1 85.00 Sulfur 1.50 wt %

C2 7.00 Inerts 0.40

C3 3.00 Water 0.10

iC4 1.50 Nitrogen 0.15

nC4 2.00

iC5 0.20

nC5 0.29

H2S 0.01

♦ Ambient air 90°F

♦ Fuel gas temperature 90°F

♦ Fuel oil temperature 175°F

♦ Atomizing steam 2.50 lbs/USG of fuel (steam at 150 psig, 500°F)

♦ Process heat required 40.00 MM BTU/hr

♦ 50% supplied by gas, 50% by oil

♦ Oxygen content of flue gas (dry basis) 6.50 wt %

♦ Temperature of flue gas exit, 700°F

Scenario 1 Base Case: The oxygen content of the flue gas is equivalent to an excessair rate of 41.7%. This is considerably above the recommended 15-25%. Also, thestack exit temperature (700°F) is extremely high, although quite common in olderheaters. The combination of high excess air and high stack exit temperature result in aheater efficiency of 78.6% (lower heating efficiency).

Scenario 2: By reducing the air flow to the burners so that the oxygen content of theflue gas falls to 3.0% (dry basis), the excess air rate will be reduced to 15.5%. This

Page 46: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 32

action, while involving no capital outlay, will increase the heater efficiency to 83.3%.Note that, by reducing the excess air flow, the stack exit temperature will fall. SeeFigure 3.2 above. For this exercise, it is assumed that the stack outlet temperature (inthis case, the outlet of the convection section) will decrease 4°F for every decrease influe gas flow of 1,000 lbs/hour. This is in line with the data presented in Figure 3.2.Because the heater is more efficient, there will be less flue gas for the same amount ofheat absorbed by the process. Therefore, the stack outlet temperature will fall toapproximately 645°F.

The improvement in heater efficiency (4.7%) means that fuel consumption is reduced by2.9 MM BTU/hr, while still maintaining 40.0 MM BTU/hr of process heat. At a fuel valueof $2.00 (US) per MM BTU, the fuel savings total over $50,000 annually.

Scenario 3: Significant energy savings will be achieved by recovering the heat lost viathe flue gas. Waste heat recovery will involve capital investment. In this scenario, it isassumed that the temperature of the stack exit gas is reduced to 350°F. In actualapplications it will be necessary to determine the acid gas dew point and then add asafety margin to account for changes in sulfur content of the fuel and variations in heattransfer in the waste heat recovery unit. See Sections 5.2 and 7.3 for further details.Based on the data presented there, a fuel oil with a sulfur content of 1.50 wt % will havea dew point of less than 300°F. Fuel gas with low H2S content will also have a dew pointless than 300°F. Therefore, the assumed stack exit temperature of 350°F should besuitable.

Reduction of the stack gas temperature to 350°F can be realized by a number of ways:

♦ By heating another stream (Scenario 3)

♦ By heating the process stream going to the heater (Scenario 4)

♦ By heating the combustion air/excess air and the fuels and the process feed to theheater (Scenario 5).

Reducing the stack temperature to 350°F, with the excess air rate held at 15.5% willraise the heater efficiency to 91.4%. The base case load of 40.0 MM BTU/hr is stillmaintained in the radiant and convection sections of the heater and an additional 3.9MM BTU/hr is recovered from the flue gas. The fuel savings are dependent upon theefficiency of the heater/boiler that originally supplied the 3.9 MM BTU/hr. Assuming thatthe efficiency of the other heater was less than the heater in these scenarios, the heatsavings are at least 3.9/0.914 = 4.27 MM BTU/hr, with an annual value of nearly $75,000(US).

Scenario 4: In this case, the waste heat in the flue gas is used to preheat the processfluid going to the heater itself. Since only 40.0 MM BTU/hr of process heat is required, itmeans that the heater duty can be reduced. The efficiency remains the same – 91.4%.The estimated fuel savings are 4.26 MM BTU/hr compared with the results in Scenario2, i.e., the same as in Scenario 3. However, the heater is partially unloaded.

Scenario 5: The third option is to use the waste heat to preheat the combustion air, thefuel gas, the fuel oil and/or the process fluid. Preheating the air and/or the fuel putsmore energy into the heater firebox, thereby leaving more heat available for transfer to

Page 47: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2933

the process. For this exercise, the fuel gas, fuel oil and the air are preheated to 300°F.The remainder of the available heat (1.0 MM BTU/hr) is used to preheat the processfluid. Again, the fuel savings, compared with Scenario 2, are 4.26 MM BTU/hr. Thedifference between this scenario and the previous two is that the adiabatic flametemperature increases from 3,309°F to 3,471°F. This has implications regardingemissions of NOx. See Chapter 9.

Implementing Scenario 3, 4 or 5 involves a capital expenditure. The choice will dependupon a number of factors:

♦ The ability to find a heat sink for the waste heat in the stack gas.

♦ The required size of the waste heat recovery equipment.

♦ The need to unload process heaters so that problems can be eliminated or unitcharge rates can be increased.

♦ The need to avoid air quality problems.

♦ The presence of potential safety issues.

Table 3.4: ENTHALPY OF FLUE GAS COMPONENTS1

BTU/lb

Enthalpy of Stack Flue Gases Outlet Conditions 200-1600°FTemperature, °F 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600SO2 16.5 34.8 53.1 71.7 90.3 109.0 127.9 145.7 165.7 185.7 205.6 225.6 245.6 265.6 285.6H2O (vapour) 58.6 104.8 151.7 198.2 245.8 294.2 343.3 393.2 443.9 495.5 547.9 601.8 655.9 711.6 766.4N2 34.8 59.7 84.7 110.0 135.4 161.1 187.1 213.3 239.8 266.6 293.6 320.9 348.4 376.3 404.4O2 30.9 53.4 76.2 99.4 122.9 146.8 171.1 195.7 220.6 245.8 271.2 296.8 322.7 348.7 374.9Air 34.0 58.4 83.1 107.9 129.5 158.4 184.1 210.2 231.8 256.2 280.5 305.4 330.2 355.5 380.7CO2 29.4 52.1 75.7 100.0 125.6 151.2 177.3 204.1 231.4 259.4 287.9 317.1 346.8 377.2 408.1

Enthalpy of Firebox Gases Firebox Conditions 2400-4400°FTemperature, °F 2400 2600 2800 3000 3200 3400 3600 3800 4000 4200 4400SO2 (est'd) 457.4 503.1 548.9 594.6 640.4 686.1 731.8 777.6 823.3 869.1 914.8H2O (vapour) 1255.3 1384.0 1514.9 1647.9 1782.8 1919.5 2058.1 2198.3 2340.2 2483.7 2628.6N2 642.4 702.0 762.2 822.9 884.0 945.6 1007.6 1070.0 1132.8 1195.9 1259.4O2 594.1 649.1 704.7 760.7 817.1 873.9 931.2 988.7 1046.7 1104.9 1163.5Air 614.9 694.4 750.5 807.4 865.5 925.0 1014.3 1070.9 1140.5 1210.9 1282.1CO2 659.3 722.4 786.3 850.8 915.9 981.7 1048.0 1114.9 1182.3 1250.2 1318.6

1 Energy Management Handbook

Page 48: Optimizacion de Combustion ARPEL
Page 49: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2935

4.0 TYPES OF FIRED EQUIPMENT

There are many designs of fired heaters and boilers. Descriptions of these lie outside the scopeof this document. The major classifications will be briefly discussed in order to provide abackground for the remaining chapters of this guideline. Similarly, there are numerous burnerdesigns. Again, only the major classifications will be presented.

4.1 Direct-Fired Heaters

Direct-fired heaters use the heat from the flue gas and/or the radiant heat from the flameto heat the process fluid. The heaters are classified primarily according to their shapewith sub-classifications according to the path followed by the flue gas and the orientationof the coils. The classifications are cylindrical, cabin and box. The cabin heater has asloping roof between the radiant and convection sections. The radiant section of a boxheater has a flat roof. The coils in cylindrical heaters are frequently vertical but aresometimes helical. The coils in box heaters can be horizontal, vertical or arbour-shaped(in the form of an arc at the top of the radiant section). The flue gas usually flowsupward but there are designs where there is downflow to the stack and others where theflue gas flows lengthwise along the radiant section. Figure 4.1 shows examples of thecabin and cylindrical classifications.

Cylindrical heaters are typically 10-15% cheaper than cabin heaters for the sameprocess duty except when the duty is less than 10 MMBTU/hr. They also require lessplot space for installation. Fewer internal tube supports and soot blowers are required.Air preheat facilities are smaller. Large cylindrical heaters have more natural draft andhigher convection heat transfer coefficients than cabin heaters. However, cabin heatershave fewer problems with two-phase flow. (Cylindrical heaters are prone to problemswith slug flow.) Cabin heaters can have upward-firing, but they can also have end andside-firing. Therefore, they can be built closer to the ground than cylindrical heaters,which must have upward firing burners. By installing a bridgewall down the center of thecabin, a cabin heater can have dual service.

The two main heat transfer compartments in a heater are the radiant section and theconvection section. Some cylindrical heaters consist of only a radiant section. They arecharacterized by low thermal efficiency. There are also heaters that have onlyconvection sections.

The radiant section is the portion of the heater that contains the burners. Heat transferis by radiation. A well-designed and operated radiant section will have no flameimpingement on the tubes. The flame length is generally 60% of the firebox and theedge of the flame should be at least 1.5 feet from the tubes. Flue gas leaving the radiantsection should be 1,500-1,900°F.

Page 50: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 36

Figure 4.1: TYPICAL DIRECT FIRED PROCESS HEATERS1

For small cabin heaters the fire box should be of equal dimensions (width, height andlength). For larger units the relative dimensions (W:H:L) are typically 1:2:4. For smallcylindrical heaters the height of the radiant section is equal to the diameter of the tube

1 GPSA Engineering Data Book, 10th Edition; page 8-15.

Page 51: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2937

circle. (See Figure 4.1). For large cylindrical heaters the length of the radiant sectionshould be twice the diameter of the tube circle.1

The convection section is downstream of the radiant section (the physical locationdepends upon the heater construction). Heat is transferred from the flue gas.Depending upon the surface area of the tubes installed there, the flue gas leaving theconvection section will have temperatures from 300-1,600°F. Older heaters tended tohave less surface area, i.e., higher flue gas exit temperatures and therefore, lowerefficiency. Plants have resorted to a number of options for increasing heat transfer inthe convection section and in the lower stack:

♦ Adding more tubes to the convection section

♦ Installing studded tubes, or even better, finned tubes in the convection section

♦ Installing waste heat recovery units into the stack.

For more detail, see Chapter 7.

4.2 Fired Boilers

Fired boilers consist of two main types: fire-tube and water-tube. In the fire-tube type,the fuel is fired in the furnace and the flue gas travels through the tubes. The furnaceand tubes are installed within a vessel that contains the steam and water. The flue gasmay travel through 2-4 passes before exiting via the stack. The boiler may have a dryback, wherein a refractory-lined chamber directs the flue gas from the furnace to thetubes. If the boiler has a wet back, the refractory is replaced by a water-cooled jacket.

A very common example of this form of boiler is the Scotch Marine boiler. It is apackaged unit, meaning that the burner, controls and auxiliary equipment are designedas a single engineered package and ready for on-site installation.

A fire tube boiler is generally limited to 350 psig steam.2 They are rated in boilerhorsepower (BHP), where 1 BHP = 33,475 BTU/hr.3 Ratings for fire tube boilers rangefrom 15-1,500 BHP (0.5 MMBTU/hr to 50 MMBTU/hr).4

Water-tube boilers are generally used for steam pressures greater than 350 psig. In thistype of boiler, the tubes are filled with water/steam, while the flue gases flow around thetubes. A steam drum is mounted at the top of the tubes and a mud drum is mountedbelow the tubes. Packaged water-tube boilers can generate steam up to 200,000 lbs/hrat 1,000 psig and 850°F.5

Common configurations for packaged water-tube boilers are the D-type and the O-type.See Figure 4.2. In the D-type, the convection tubes may contain a superheater. It is

1 GPSA Engineering Data Book, 10th Edition; page 8-15.2 Cleaver-Brooks; http://www.cleaver-brooks.com/Boilersa/.html and /GlossFP.html3 Ibid, http:/www.cleaver-brooks.com/GlossAE.html4 Ibid, http:/www.cleaver-brooks.com/Boilersa/.html5 V. Ganapathy, Understand Boiler Performance Characteristics; Hydrocarbon Processing; August 1994;

page 131.

Page 52: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 38

installed in the convection section at a point where the flue gas is hot enough to deliverthe desired steam temperature. If only 20-50°F of superheat is required, thesuperheater is usually mounted between the evaporator and the economizer. The fluegas leaving the convection section is used to preheat boiler feedwater in an economizer.Air preheaters on packaged boilers are not common due to cost considerations, largegas/air pressure drops and increased formation of NOx.

1

The O-type boiler is not as easy to fit with an economizer or superheater. The design issymmetrical so the least tube surface is exposed to radiant heat.2 The A-type has twosmaller lower drums or headers and a large upper drum for steam and waterseparation.3

Packaged boilers usually have forced-draft air supply. This allows pollution abatementequipment such as flue gas recirculation, staged-fuel or staged-air burners. Pressuresin the furnace can reach 10-30 inches of H2O. This could lead to emissions of carbonmonoxide if the partition wall, which separates the furnace from the convection section,develops leaks.4

It is very important to match the design of a packaged boiler with its normal operatingconditions. As the load on the boiler is decreased, the steam temperature, the flue gasexit temperature and the boiler feedwater temperature leaving the economizer alsodecrease. This phenomenon is discussed in Section 3.2. For the effect on the operationof a packaged boiler, see Figure 4.3.

Note that the efficiency of the boiler has a maximum value. This is the result of threefactors. The flue gas temperature increases with increasing load, and as seen inChapter 3, this leads to a loss of thermal efficiency as more heat is escaping out thestack. Another loss is due to radiation and convective losses. Packaged boilerstypically have radiation and convective losses of 0.5% of the heat released at full load.5

Since the surface temperature of the boiler remains relatively constant over theoperating range, the heat loss in quantitative terms is also relatively constant. However,in terms of the percentage of heat release, the radiation loss increases to approximately2% at 25% load. Moreover, in order to maintain good firing conditions the excess airmay have to be increased at low boiler loads.

1 V. Ganapathy, Understand Boiler Performance Characteristics; Hydrocarbon Processing; August 1994;

page 131.2 J. Makansi; Managing Steam; page 19.3 Ibid.4 V. Ganapathy; Hydrocarbon Processing; August 1994; page 132.5 Ibid, page 132 and H. Hendry (John Inglis Equipment); Boiler Efficiency and Testing; Publication B623;

no date.

Page 53: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2939

Figure 4.2: COMMON BOILER CONFIGURATIONS1

1 V. Ganapathy; Hydrocarbon Processing; August 1994; page 132.

Page 54: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 40

Figure 4.3: PERFORMANCE OF A PACKAGED BOILER1

200

300

400

500

600

700

800

20 30 40 50 60 70 80 90 100

Percent Load or Steam Flow

Tem

per

atu

re, °

F

81

82

83

84

85

86

87

Eff

icie

ncy

(H

HV

), %

Steam Temperature

Efficiency

Boiler Feedwater Leaving Economizer

Flue Gas Exit Temperature

4.3 Incinerators

4.3.1 Incinerators for Process Wastes

Incinerators are used to destroy hazardous wastes and pollutant gases such as sulfurplant tail gas. Depending upon the heating value and quantity of the waste,supplemental fuel may be used. There are burners capable of burning three fuels: oil,gas and low-BTU waste gas. There is a tip for each fuel, which can be burnedsimultaneously. The single most important factor that determines whether supplementalfiring is required is the hydrogen content of the waste gas. Normally, if the fuel contains10-15 volume % hydrogen and the heating value of the gas is at least 80-100 BTU/ft3, nosupplemental fuel is required. However, a gas containing 25% methane and 75% CO2

has a heating value of 227 BTU/ft3 but will require either supplemental firing or a specialcombustion chamber.2

Descriptions of incinerators for waste solids and liquids are provided in the ARPELGuidelines for the Management of Petroleum Refinery Solid Wastes (Section 5.5.3) andGuidelines for the Management of Petroleum Refinery Liquid Wastes (Section 5.5.3).

Incineration of the tail gas from sulfur recovery units is very common. The oxygen in theflue gas is often in the range of 5%. Use of high intensity burners can significantlyreduce supplemental fuel consumption by providing better air-fuel mixing. This results ina shorter, hotter flame, which ensures more complete destruction of the pollutants (H2S,COS, CS2 and S). Reduction of the oxygen in the flue gas to as low as 2-3% is possible,resulting in fuel savings of up to 30%. However, it is important to review the impact ofthese changes upon downwind air dispersion. With less flue gas going up the stack the

1 V. Ganapathy; Hydrocarbon Processing; August 1994; page 135.2 John Zink Company; Combustion and Industrial Burner Application and Design; no date; page 42-44.

Page 55: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2941

plume rise and subsequent dispersion of the SO2 will be altered. It is necessary toconfirm that the revised incinerator operations still allow compliance with applicable airregulations.

4.3.2 Incinerators for Medical Wastes

A number of ARPEL facilities also support non-industrial activities such as hospitals.Incineration of medical wastes is therefore, necessary. Wherever possible, they shouldbe sent to an incinerator licensed to handle them. If the incineration is done on-site, acontrolled air incinerator, with primary and secondary combustion chambers and aheated hearth is necessary. See Sections 5.5.3.3, 5.5.3.3.1 and 5.5.3.3.2 of the ARPELGuidelines for the Management of Petroleum Refinery Solid Wastes for furtherinformation.

4.4 Other Heaters

Among the less-common types of fired heaters found in the oil and gas industry are thefollowing:

♦ Convection heaters. These have no radiant section. They can be equipped withflue gas recirculation and pre-mix burners. As a result, the heaters have greater gasflow (higher heat transfer coefficients) and about 10% excess air. These combine toproduce a highly efficient heater. Convection heaters also offer a high degree ofsafety and are therefore ideal for installations such as offshore platforms.

♦ Bath heaters. This is an indirect fired heater. The fuel is combusted in the fire tubeand the heat is transferred to the heat medium, which in turn transfers the heat to theprocess fluid. The desired bath temperature determines the heat medium. SeeTable 4.1.

A common use for bath heaters, especially in the production sector of the oil and gasindustry, is for providing heat to circulating heat media such as glycol. The heatmedia is often used in low-temperature reboilers to separate LPG from solution gasand natural gas for heating buildings.

Other uses include the heating of natural gas pipelines from the wellhead to theproduction facility in order to keep the gas above the hydrate point and to heatoil/water emulsions in order to provide better separation.

Page 56: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 42

Table 4.1: TYPICAL BATH HEATERS1

Heat Medium Bath Temperature, °F Thermal Efficiency (LHV), %

Water 180-195 76-82

50% Ethylene Glycol 195-205 76-80

Low Pressure Steam 245-250 76-80

Hot Oil 300-550 71-76

Molten Salt 400-800 68-74

TEG Reboiler 350-400 75-80

Amine Reboiler 245-270 75-80

4.5 Burners

Brief descriptions of the various types of burners commonly used in the oil and gasindustry are provided. For information regarding the construction and operation of theseburners, the reader is referred to the article Combustion and Industrial BurnerApplication and Design issued by John Zink Company.

The burners discussed below are generic types. They have been included to providethe reader with the range of options available and the effects that burner type has oncombustion. The wide variety of burner models, even within each generic type,precludes descriptions of individual brand names. Each facility should consult theequipment vendors and manufacturers for their expertise concerning site-specific burnerissues.

♦ Premix burners. The primary air is pulled into the fuel gas by means of a venturi.Secondary air registers supply the remaining combustion air. The flame is short anddense. The normal mode of operation is to have the primary air registers fully openand the secondary air adjusted to the desired excess air. (The primary air registersare adjusted if there is flashback or flame liftoff.) This burner operates on a naturaldraft heater.

♦ Nozzle-mixing gas burners. The air and gas are kept separate until they leave thenozzle. These are also known as raw gas burners. They have excellent turndownratios (10 to 1), whereas premix burners have a nominal turndown ratio of 3 to 1.Although the burners require lower gas pressures than premix burners, there stillmust be sufficient pressure to ensure good air-fuel mixing if low excess air operationis desired. All in all, the mixing is not as good as with premix burners. Nozzle-mixinggas burners also operate with natural draft air supply.

♦ Combination natural draft gas and oil burners. Most oil burners in refineryoperations have the capability of firing gaseous fuels. The gas firing ports providethe flame pattern. The position of the oil tip is critical. If it is too high there may be

1 GPSA Engineering Data Book, 10th Edition; page 8-25.

Page 57: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2943

flame instability and if it is too low there may be coking on the burner tile and an oilspill. The operation of simultaneous oil and gas firing is discussed in Section 5.3.

♦ Low air pressure drop burners. These are forced draft burners and are normallyused when low air pressure drops are available or when there is the choice ofoperating in a natural draft mode. The natural draft mode determines the size of theburner but the materials of construction are designed for preheated air temperatures.

♦ High air pressure drop burners. This is a forced draft, dual-fuel burner that is wellsuited for heaters with a single burner. Rotating vanes in the air register impart aspinning motion to the air, which causes a short, wide flame. There is better mixingthan with the low air pressure burner, which allows a lower gas pressure (5 psigversus 15 psig) and possibly less atomizing steam. Large burners have a dual zoneair register to ensure that the oil receives sufficient air. There are variations of thisburner.

♦ High intensity burners. These require forced-draft air supply. The air passesthrough a series of spin vanes, causing a vortex. Most, (75+%) if not all, of thecombustion occurs in the refractory lined combustion chamber. The flame is verycompact. Because the heat is retained in the chamber makes it an ideal burner forheavy oils and gases with low heating values.

Table 4.2 summarizes the specifications of the above-mentioned burners. This isintended as a guide. Burner manufacturers should be consulted for their input.

Table 4.2: BURNER SPECIFICATIONS1

Burner Type Air Draft Fuel Heat Release Normal Draft Normal Minimum Flame Flame FlameMMBTU/hr inches WC Excess Air, % Shape Length Diameter

Premix Natural Gas 0.1 - 15.0 <0.1 - 1 5 Conical Various VariousFlatFan

Round FlatNozzle-Mixing Natural Gas 0.5 - 20.0 0.1 - 1 5 - 10 Conical Various Various

FlatFan

Combination Natural Gas 1.0 - 20.0 0.1 - 1 5 - 10 for gas Conical Various VariousOil 10 - 15 for oil Flat

FanLow Air ∆P Forced Gas 1.0 - 30.0 2+ 5 - 10 for gas Conical Various Various

Oil 10 - 15 for oil FlatFan

High Air ∆P #1 Forced Gas 15.0 - 200.0 2 - 10 5 Conical Various VariousOil

High Air ∆P #2 Forced Gas 5.0 - 40.0 1 - 9 <5 Conical Various VariousOil

High Intensity Forced Gas 5.0 - 100.0 6 <5 CompactOil

Considerable attention has been paid to burners in Claus Sulfur Recovery Units becauseof the need to maintain very close combustion control and the need to ensure thermaldestruction of the tail gas. For further information the reader is referred to theJanuary/February 1993 issue of the magazine Sulfur, for the article Leading BurnerDesigns for Sulfur Plants, pages 23-34.

1 John Zink Company; Combustion and Industrial Burner Application and Design; no date; pages 10-40.

Page 58: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 44

Low-NOx burners are becoming very important in light of the increasingly stringent airquality regulations. Among the options for reducing NOx is the use of staged-air andstaged-fuel burners. These are discussed in Section 9.3.

4.6 Design Criteria

This chapter has provided an indication of the complexity involved in designing firedheaters and boilers. The same complexity exists when specifying burners andperipheral equipment for improving thermal efficiency and complying with air qualityregulations. It is recommended that heater, boiler and burner manufacturers be includedinto the project teams investigating either new heaters/boilers or ones undergoingmodifications and upgrades. These specialists should be contacted as early as possiblein the study, as their input will significantly influence the final design.

It is incumbent upon the facilities to have sufficient high quality data of past performancethat modifications and upgrades can be designed and easily incorporated into theexisting equipment. It is also essential that the plant have as much information aboutfuture operation as possible for both upgrades and new-fired heaters/boilers. The entirerange of operating conditions should be clearly stated. Existing and expectedenvironmental regulations must also be known.

The simultaneous, and at times opposing, demands of energy conservation and airquality compliance have reduced the operating region available to staff. Tighter controlof fuel and airflow to the heaters is a must if one is to meet these demands whilemaintaining safe operation. The reader is referred to the article Controlling FiredHeaters by W. Driedger in the April 1997 issue of Hydrocarbon Processing, pages 103-118.

Page 59: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2945

5.0 MONITORING PERFORMANCE

Good monitoring of performance and proper operating practices can achieve significantefficiency in heaters and boilers. Options requiring capital outlay are discussed in Chapters 9-11.

5.1 Fuel Analysis

Analysis of the fuel(s) is necessary to know the heating value, the required air and theresulting adiabatic flame temperature. Also, contaminants such as sulfur/H2S andnitrogen will determine the emissions of SOx and fuel NOx. These in turn dictate theminimum flue gas temperatures that can be achieved in waste heat recovery units.

It is recommended that regular testing of fuel(s) consumed at the facility be conducted.This is especially necessary if fuel gas is being supplied by processes subject tocatalytic deactivation. For example, catalytic naphtha reformers produce large quantitiesof hydrogen at the beginning of their cycles. If the unit offgas is sent to fuel, the quantityof hydrogen will have an impact on the fuel characteristics. If the facility imports its fuel,it should ask the supplier for analyses on a regular basis.

The term “regular testing” is defined as being frequent enough to detect normal changesin fuel quality. This will vary from facility to facility and will have to be determined byexperience. At least once per week for refineries and gas plants is suggested. Standardfuel testing protocols, as specified by ASTM or equivalent, should be followed.

The analysis of a fuel must be sufficient to allow a determination of its heating value andits contaminants. For gaseous fuels, this will consist of a composition at least as far asC6-C8. For liquid fuels, density values and sulfur levels are required frequently andanalyses for inert materials, water and nitrogen less frequently.

5.1.1 Wobbe Index

When fuel quality changes, the plant operators must consider the following:

♦ The same heat input rate must be maintained for similar process conditions.

♦ The heater/boiler equipment must be able to handle the new fuel. Equipmentincludes the stack, burners, piping valves and controls.

♦ The stability of the burner, the heat release pattern and the furnace atmosphere mustremain constant.

The Wobbe Index enables the operators to evaluate changes to the first two pointsmentioned above. The index is confined to gas-gas comparison. If the old and newgases have the same index value, there is no need to change valve settings in the fuelgas supply piping.

Page 60: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 46

The Wobbe index is given by the formula:1

Wo = Ho/(Go)0.5 = Hn/(Gn)

0.5

where:

Wo = Wobbe Index

Ho = heating value of the old fuel

Hn = heating value of the new fuel

Go = gravity of the old fuel, relative to air

Gn = gravity of the new fuel, relative to air

One company, faced with widely and quickly varying fuel gas quality used the output of aWobbe Index meter, regression analysis and feed forward control to operate their utilityboilers at 1.5-2.0% excess air, or about 0.5% oxygen in the stack.2 The regressionanalysis was used to correlate to correlate the Wobbe Index with fuel heating value,specific gravity and air requirements. Note that many factors such as fuel quality, burnertype, air supply and air quality emission levels must be considered when operating atsuch low excess air levels. However, with appropriate data and control schemes, it maybe possible to achieve those levels.

Feedforward air control is especially useful if the heater or boiler is firing, as its entire orpartial fuel, waste gases since they are subject to wide variation in composition and flow.This situation is discussed in the article Feedforward Air Control for Fuel BTU Changesby E. Vicknair, in the July 1985 issue of Hydrocarbon Processing, pages 65-66.

5.2 Stack Gas Monitoring

The stack gas from each heater and boiler should be tested as frequently as possible.Many heaters have oxygen analyzers and/or controls on the stack. These should bechecked on a regular basis. As discussed in Section 5.3, the greater the draft (the largerthe negative pressure) the greater the ingress of air through leaks and other openings inthe heater. For this reason, it is suggested that the flue gas be sampled as close to thefirebox as possible. (The highest pressure (lowest draft) is recorded at the arch, or entryinto the convection section. If the sample is taken at the arch, the probe should bezirconium oxide as the temperature is typically in the range of 1,400-1,800°F.) Avoidtaking the sample near or downstream of an induced draft fan as the suction of the fanhas a very low pressure (i.e., potentially high ingress rates of air).

Flue gas analyses can be done using a variety of equipment such as gaschromatographs, ORSAT apparatus, and Bacharach (Fyrite) analyzers. These devices

1 North American Combustion Handbook; page 39.2 T. Mort and I. Verhappen; Improving Fuel Efficiency with Statistics; Chemical Engineering; June 1991;

pages 143-146.

Page 61: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2947

vary in sophistication. Some are capable of measuring only one gas – some Fyriteanalyzers measure oxygen and others measure carbon dioxide – while others, such asthe ORSAT apparatus can measure oxygen, carbon monoxide and carbon dioxide.Whatever device is used, its accuracy must be confirmed using test standards on aregular basis. As discussed later in this chapter, air leakage into the heater is a commonproblem. This leakage will provide falsely high readings of the oxygen in the flue gas.This becomes especially critical when operating at very low excess air rates or if the airsupply (by a fan) is controlled by the oxygen content in the flue gas. When the amountof combustion air falls below the stoichiometric amount, the generation of hydrogen andcarbon monoxide rapidly escalates. See Figure 5.1. To protect the heater/boiler froman unsafe situation, analyzers for combustible gases (hydrocarbons, hydrogen andcarbon monoxide) are also installed in the flue gas stack. The reading of this instrumentis used to override the reading on the oxygen analyzer.

It is suggested that the efficiency of each heater and boiler be calculated at least weekly.This can be done using tables, graphs, nomographs or spreadsheets. For a betterunderstanding of the heater performance, the efficiency should be tracked over time.Key operating parameters should be noted so that the causes of swings in efficiency canbe determined. Readings of the oxygen in the stack gas and the stack gas exittemperature should be reported at least daily and preferably each shift.

Heater/boiler efficiency is primarily a function of the excess air rate and the flue gas exittemperature. Other factors include the ambient air temperature and the fuel conditions.It is recommended that the monitoring of heater performance include the simultaneousrecording of the flue gas analysis and the stack exit temperature, as well as the ambientconditions. It is preferable that the fuel analysis and fuel temperature be recorded atroughly the same time as the stack monitoring.

Another important variable to monitor is the draft profile throughout the heater and duct.This is especially necessary if the heater/boiler is suspected of poor operation. Whenconducting a draft survey, record the damper position and the condition of anyequipment in the stack downstream of the convection section. Ensure that the draftgauges are operating properly and that the instrument lines are clear.

More intensive surveys such as thermography and analyses of the gases within thefirebox and convection sections are recommended if the heater is operating inefficientlyand normal practices (see Section 5.3) are not producing the desired effect. Thesurveys are also required if heater capacity upgrades are planned.

5.3 Firing Practices

The following procedures should be done to optimize heater/boiler operation.

Reduce excess air to the levels specified by the burner manufacturer. These levels willvary depending upon the type of burner, the type of air supply (natural draft or forceddraft) and the type of fuel. Recommended excess air levels are provided in Table 5.1.

Page 62: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 48

Table 5.1: RECOMMENDED EXCESS AIR LEVELS,1 %

Natural Draft Forced Draft

Fuel Gas 15-20 10-15

Light Fuel Oil 20-25 15-20

Heavy Fuel Oil 25-30 20-25

Table 5.1 should only be used as a guide if the manufacturer’s recommended excess airrates are not available. The values in the table presume the heaters are in goodcondition with minimum air leaks. Moreover, low-NOx burners tend to require moreexcess air.

It is extremely important not to reduce excess air rates to the point where flameinstability or incomplete combustion occurs. Theoretically, the stoichiometric amount ofair would result in zero percent oxygen in the flue gas. However, air leaks into theheater and stack and poor air-fuel mixing lead to the situation where a “measured” 5-15% excess air is needed to achieve “stoichiometric” conditions at the burner.

Airflow to the burner is varied by controlling the draft of the heater. The available draft ina heater is given by the equation:2

Draft = 0.192 * Hs * (ρg – ρa) + 0.0029 * V2 * ρg * (4 * f * Hs/D+1)

where

Draft = available draft, in inches of water

Hs = height of stack, feet

ρg = density of flue gas, lbs/cubic foot

ρa = density of air, lbs/cubic foot

V = velocity of flue gas, feet/second

f = Fanning friction factor

D = diameter of stack, feet

The Fanning friction factor can be found in a variety of books, such as Perry’s ChemicalEngineers’ Handbook, 4th Edition, page 5-20.

1 A. Garg; Optimize Fired Heater Operations to Save Money; Hydrocarbon Processing; June 1997; page

103.2 GPSA, Engineering Data Book,10th Edition; page 8-10.

Page 63: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2949

For heaters that are natural draft or forced-draft, the draft is controlled by adjusting thedamper in the stack. For heaters that are equipped with an induced-draft fan on theoutlet of the convection section, the draft is controlled by the dampers on the suction ofthe induced-draft fan. The draft can also be controlled by the air registers at the burners.

The draft equation noted above consists of two terms. The first term is always negativebecause the density of the flue gas is always less than the density of the air. Thisprovides the total available driving force of the gas. The second term accounts for thepressure drop due to friction and the stack exit velocity.1 Typically, there is a pressureincrease in the radiant section of 0.01 inches of water per foot of radiant section height.2

The highest pressure in the heater is at the arch, or top of the radiant section.3 Thisvalue is used to control the heater draft.

The higher the draft, the greater the amount of air drawn into the heater. Conversely,the lower the draft, the smaller the amount of air drawn into the heater. If the draft is toolow, unsafe conditions could exist, wherein flue gas would leak out of the heater. Tomaintain proper draft control, a pressure reading of –0.1 inches of water should bemaintained at the arch.4 This results in safe operation and minimizes air leaks.

Closing the stack damper reduces the draft (i.e., the pressure at the arch moves closerto zero). This changes the draft profile across the entire heater. Opening the airregisters also reduces the draft but only as far as the damper. To adjust the quantity ofexcess air, both the damper and the air registers must be adjusted.5

The API Document 535 titled “Burners for Fired Heaters in General Refinery Service”, 1st

Edition (July 1995) published a flow chart for adjusting the draft and oxygen in the fluegas. The contents are summarized here:6

♦ If the draft at the arch is at the target level, check the oxygen in the flue gas. If theO2 is on target, there is good operation. If the 02 is higher than target, close thedamper. If the O2 is below the target, open the air registers.

♦ If the draft at the arch is higher than the target level (i.e., greater negative pressure),check the oxygen in the flue gas. If the O2 is on target, open the air registers andclose the damper. If the O2 is higher than target, close the damper. If the O2 isbelow target, open the air registers.

♦ If the draft at the arch is lower than the target (i.e., smaller negative pressure), checkthe oxygen in the flue gas. If the O2 is on target, open the damper and close the airregisters. If the O2 is higher than target, close the air registers. If the O2 is belowtarget open the damper.

Figure 5.1 shows a typical draft profile for a direct fired heater. Note that is slightlydifferent from the profile printed in API Document 535, 1st Edition (July 1995) because it

1 GPSA Engineering Data Book, 10th Edition; page 8-10.2 Ibid, page 8-18.3 A. Garg; Hydrocarbon Processing; June 1997; page 99.4 Ibid and discussions with John Zink Canada.5 A. Garg; Hydrocarbon Processing; June 1997; page 101.6 Ibid.

Page 64: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 50

has combined the pressure drop and the stack effect across the convection section andit has included the pressure drop across the burners.

Figure 5.1: TYPICAL DRAFT PROFILE IN A DIRECT FIRED HEATER1

Stack

Damper

Convection Section

Radiant Section

Burners

ooooooooooooooo

ooooooooo

ooooooooo

Breeching

Arch

oo oo

NegativePressure

Positive Pressure

Maintain at -0.1 inches of water column

If the amount of air into the firebox drops below the stoichiometric requirement, there willbe a rapid increase in the amount of carbon monoxide and hydrogen produced. SeeFigure 5.2. (Note, this is the basis of staged burners. See Section 9.3.)

Figure 5.2: EMISSIONS OF HYDROGEN AND CARBON MONOXIDE UNDER SUBSTOICHIOMETRICCONDITIONS2

0

1

2

3

4

5

6

7

8

9

60 70 80 90 100 110

% of Stoichiometric Air

Vo

lum

e %

(w

et b

asis

)

Theoretical Equilibrium Concentration of H2

Theoretical Equilibrium Concentration of CO

H2

CO

Fuel is Methane

1 GPSA Engineering Data Book; 10th Edition; page 8-19.2 Burner Design Parameters for Flue Gas NOx Control; by R. Martin, John Zink Company; no date;

Figures 1 and 2.

Page 65: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2951

Assuming that the heater is being operated with proper draft control and with sufficientair to achieve good flame shape and complete combustion with the available burners,the efficiency of the heater can be calculated using the procedures and data presentedin Chapters 1-3. A quick estimate for heaters firing natural gas can be obtained fromFigure 5.3.

Figure 5.3: THERMAL EFFICIENCY (LHV) FOR A HEATER FIRING NATURAL GAS1

65

70

75

80

85

90

95

300 400 500 600 700 800 900 1000 1100

Stack Exit Temperature, °F

Th

erm

al E

ffic

ien

cy (

LH

V),

%

15%

50%40%30%

20%

25%

Parameter is Excess Air

If the heater conditions lie outside the range of the parameters in Figure 5.3, theefficiency can be estimated using Figure 3.1. See Figure 2.4 for the relationshipbetween oxygen in the flue gas and excess air rate. Note that Figure 5.3 expressesefficiency in terms of its lower heating value, where as Figure 3.1 is in terms of its highheating value. It is best to work entirely on one graph or the other and to note the basisof the efficiency calculation (LHV or HHV). For plants that measure the carbon dioxidein the flue gas, instead of the oxygen, Figure 5.4 can be used if the heaters burn naturalgas and Figure 5.5 if they burn No. 6 Oil. Note that the graphs show the stack loss. Toconvert to efficiency, subtract the stack loss plus radiation losses from 100%.

For additional information regarding optimization of heater operation, the reader isreferred to the article Optimize Fired Heater Operations to Save Money by A. Garg in theJune 1997 issue of Hydrocarbon Processing, pages 97-104.

When firing oil, ensure that there is sufficient amount of atomizing steam to ensure goodcombustion. Not only does the steam break up the oil into droplets, it providesdischarge velocity to ensure good air-fuel mixing. Dry steam should be used: wet steamcauses “sparking”. The atomizing steam pressure should be 20-30 psi above the fuel oilpressure. Steam consumption is about 0.3 pounds of steam per pound of fuel.

1 Adapted from A. Garg; Hydrocarbon Processing; June 1997; page 101. Some points have been

adjusted to smooth the lines.

Page 66: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 52

Figure 5.4: ENERGY LOSSES WHEN BURNING NATURAL GAS1

0

10

20

30

40

50

60

200 300 400 500 600 700 800 900 1000 1100

Difference Between Flue Gas Exit and Ambient Temperatures, °F

Sta

ck L

oss

, %

3% 4%

6%

8%

10%

12%

Radiation and Convection Losses Excluded

Parameter is % CO2 in Flue Gas

Gas Heating Value 1000 BTU/SCF

Figure 5.5: ENERGY LOSSES WHEN BURNING NO. 6 OIL2

0

10

20

30

40

50

60

70

80

200 300 400 500 600 700 800 900 1000 1100

Difference Between Flue Gas Exit and Ambient Temperatures, °F

Sta

ck L

oss

es, %

3% 4%

6%

8%

10%12%

14%16%

Parameter is % CO2 in Flue Gas

Radiation and Convection Loss Excluded

Steam-assist burners use less steam. However, the fuel orifice is smaller so it is moreprone to plug and high oil and steam pressures are required.

1 Garcia-Borras, page 27.2 Ibid, page 26.

Page 67: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2953

Air atomization burners use low pressure air (1-2 psig). With a good design, 10% of thecombustion air is injected into the heater as atomizing air. Low fuel pressure can beused but the higher the oil pressure, the lower the amount of air.

Mechanical atomizers use extremely high fuel oil pressure to break up the oil. Very highpressures are obtained by having very small orifices at the fuel tip. These devices areusually only found in extremely high heat release burners.

The higher the fuel oil pressure at maximum firing the higher the available pressure atturn-down conditions. This results in better atomization even with reduced rates ofatomizing steam.1

When firing oil and gas in a single burner, the flame length is always longer, even whenthere is sufficient air. It is not possible to determine whether the proper fuel-to-air ratio isbeing maintained. It is suggested that, if two fuels are necessary, fire some burners onoil as the base-load and control the heater by firing gas on the other burners. If theflames on the oil burners are long and smoky, the burner is probably overfired and thefuel oil flow should be adjusted to get good flame quality.

If dual-firing on burners must be practiced, all burners should be base-loaded equallywith one fuel and controlled by adjusting the other fuel.2

For proper heat distribution in the firebox and burning a single fuel, maintain equal fuelpressure to each burner. Open the air registers of all burners the same amount on anatural draft heater. For heaters with forced-draft air supply, open fully the burner airdampers and control the air flow by using the fan suction damper. If the heater hasstaged-air low-NOx burners, ensure that all dampers are open the same amount.3

In order to maximize heat transfer across the available surface area, it is necessary toensure that the tubes are clean. The problem of flame impingement causing internalcoking has been discussed elsewhere in this document. External fouling by soot alsorapidly leads to poor heat transfer. See Figure 5-6. To remove soot from the convectionsection tubes, soot blowers are installed if the heaters burn oil. They are typicallyinstalled every four or five rows.4

The soot is usually removed by high pressure steam directed onto the tubes. Sootblowing is generally done only once per shift. These devices require considerablemaintenance. Among techniques to improve soot blowing is the use of microprocessorbased control that directs the soot blowing to areas of poorest heat transfer. Sonicsootblowing uses a nearly continuous sound pressure wave to keep ash and sootsuspended in the flue gas. These devices have low maintenance, are easily installedand can be fully automated. However, they can only remove light, friable deposits ofslag and soot. Vanadium and sodium in the fuel oil form sticky deposits because of theirlow melting points. Neither steam nor sonic soot blowing can easily remove suchdeposits.5

1 John Zink Company; Combustion and Industrial Burner Application and Design; no date; pages 21-27.2 Ibid; pages 27-28.3 A. Garg; Hydrocarbon Processing; June 1997; page 1024 GPSA Engineering Data Book, 10th Edition; page 8-17.5 A. Garg and H. Ghosh; Chemical Engineering; October 1990; page 222.

Page 68: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 54

Figure 5.6: EFFECT OF SOOT DEPOSITS ON FUEL COMBUSTION1

0

1

2

3

4

5

6

7

8

9

0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14

Thickness of Soot, inches

Incr

ease

in F

uel

Co

nsu

mp

tio

n, %

Scale on the heater tubes has the same effect as soot – an increase in fuel consumptionfor the same heat transfer. See Figure 5.7. Note that the type of scale is a factor.Scales containing silica are particularly troublesome. The tubes should be cleaned ofscale during shutdowns in order to improve heat transfer and heater efficiency. (Alsoduring shutdowns, the condition of finned and studded tubes should be noted andburned-off and broken fins/studs repaired or replaced.

The quality of the heater refractory and insulation should be checked frequently andrepairs made as soon as possible. Typically radiation and convective losses, due towind, are in the range of 2-3% of the heat release. They can be more accuratelyestimated using Figure 3.3.

The impact of excess air on heater efficiency has been discussed previously. Theingress of air into the heater can take place via unused burners, open sight doors,explosion doors, draft sample tubes and leaks. This air not only reduces efficiency byabsorbing heat but also creates a falsely high oxygen reading in the flue gas. SeeFigure 5.8 for an estimation of air leakage. Note that if the heater has an induced draftfan, the pressure may be as low as –10 inches of water near the suction of the fan. Anyleaks in this region will have very high leakage rates.2 Operators should ensure that allopen ports are closed if not needed and that the air registers of unused burners are shut.(An exception to this is staged air combustion. See Section 9.3.) Leaks should berepaired as soon as possible.

1 Garcia-Borras, page 41.2 A. Garg; Hydrocarbon Processing; June 1997; page 101.

Page 69: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2955

Figure 5.7: EFFECT OF SCALE DEPOSITS ON FUEL COMBUSTION1

0

1

2

3

4

5

6

7

8

9

0 0.01 0.02 0.03 0.04 0.05 0.06

Thickness of Scale, inches

Iron and Silica

High Iron Content

Normal Scale

Figure 5.8: AIR LEAKS INTO HEATERS2

0

100

200

300

400

500

600

700

-0.40 -0.35 -0.30 -0.25 -0.20 -0.15 -0.10 -0.05 0.00

Furnace Pressure, inches water column

Air

Infi

ltra

tio

n, f

t3 /ho

ur/

in2

1 A. Garg; Hydrocarbon Processing; June 1997; page 101.2 Garcia-Borras, page 54.

Page 70: Optimizacion de Combustion ARPEL
Page 71: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2957

6.0 ENERGY REDUCTION TECHNIQUES

This chapter raises the issue of energy effectiveness. This is related to, but yet quite distinctfrom, energy efficiency. As an example of these concepts, consider the generation of steam tooperate non-condensing turbines. At their best, a boiler will have an efficiency of about 90%and a turbine will be about 75% efficient. Assume 100,000 lbs/hr of 600 psia steam at 700°Fdrives the above-mentioned turbine and exits as 100 psia steam. The turbine exhaust steam isthen condensed in aerial coolers for recycling back to the boiler.

Assuming that the deaerator operates at atmospheric pressure, the fuel input to the boiler is 130MM BTU/hr. Only 17.5 MM BTU/hr is consumed in the turbine, where 13 MM BTU/hr of usefulwork is produced. The remaining energy in the steam – 100 MM BTU/hr – is dissipated in theaerial coolers (including the energy required to operate the fan motors).

In summary, only 10% of the energy input is recovered in the form of useful work. Over 76% ofthe available energy is wasted in the aerial coolers. While the individual pieces of equipmentare very efficient, the overall effectiveness of the system is very low. The object of this chapteris to discuss methods of increasing the effectiveness of heater/boiler systems.

6.1 Minimizing Heater Duties

There are two sides involved with heaters and boilers: the process side and the fuelside. This chapter deals with the minimization of heater duties using technologies andtechniques that reduce the amount of process heat that must be absorbed. Chapter 7discusses methods of ensuring that the amount of fuel input required to achieve therequired process heat is as low as possible.

The reader is urged to consult the ARPEL Guidelines for the Development of EnergyAudits in Upstream and Downstream Oil and Gas Facilities. That document provides asystematic approach for evaluating the impact of process conditions on energy input.

6.1.1 Heat Exchange

Heat exchangers are important pieces of heat transfer equipment. For many processunits, the quantity of heat transferred in exchangers is equivalent to, or even greater,than that supplied in heaters. Proper placement and operation of them can significantlyreduce the amount of heat that must be supplied by fired heaters. In evaluating theeffectiveness of heat transfer by exchangers in a plant, consider the following:

♦ The amount of heat transfer using exchangers. Are the expected temperaturesbeing achieved? If not, are the exchangers fouled? Has surface area beendecreased due to plugged tubes? Are the fins on tubes with extended surfacesdamaged? Do the exchangers have the appropriate internal configuration (i.e., theoptimum number of shell/tube passes, correct baffle spacing, tube pitch, etc.).

♦ The heat exchanger arrangement in exchanger trains. An exchanger train is a heatexchanger network wherein one common stream (e.g., such as crude oil) is heatedby a number of product streams (e.g., the products from a crude distillation unit). Tomaximize heat transfer, it is desirable to have the common stream heated by

Page 72: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 58

increasingly hotter streams. The original design usually takes this into account.However, changing product slates, changing feed quality and latermodifications/expansions to the exchanger train can combine to reduce theeffectiveness of the train.

♦ The heat exchanger arrangement in exchanger systems. By this is meant smallerexchanger networks where one stream exchanges heat with only one other. Morethan one exchanger is involved. It could also describe a subsection of an exchangertrain.

Multiple exchangers are installed in series or in parallel. Heat transfer isincreased by installing in series. However, pressure drop increases significantly.On the other hand, when installed in parallel, the plant has the ability to take anexchanger out of service for cleaning or repair with less disruption to operations.

♦ The amount of heat removed from the process by cooling water or air. Water and airfor cooling should be utilized only when the product streams can no longer feasiblysupply heat to other streams.

♦ The use of intermediate storage. In order to prevent excessive evaporative lossesand/or to maintain the integrity of the storage tanks, intermediate products from UnitA are usually cooled prior to storage. The heat exchangers in the downstream unit(Unit B) are therefore, required to raise the temperature of the incoming feed, oftenessentially back to the levels originally experienced in Unit A prior to cooling forstorage. Minimization of intermediate storage would make more process heatavailable for increasing the temperature of the feed to the heater(s) in Unit B and/orproviding heat to other locations.

♦ The amount of inter-unit heat transfer. Heat transfer between process units can bevery effective. However, the improvement in energy use must be balanced againstthe risk of upsets in one unit causing problems in other units.

Heat exchanger simulation programs are useful in evaluating individual exchangers andsimple arrangements. Technologies such as PINCH are recommended for analyzingcomplex arrangements and heat exchanger trains. It must be stated that monitoring andanalyzing a heat exchanger network can be complicated. The reader is referred to thearticle Challenges in Simulating Heat Exchanger Networks, by R. Sigal, in the October,1996 issue of Hydrocarbon Processing, pages 125-132. Nevertheless experience hasshown that significant improvements in heat exchanger performance are possible usingthese analytical tools.

6.1.2 Effective Distillation

The nature of crude oil, and to a smaller extent, natural gas, is such that distillation isrequired to manufacture useful products. Proper and effective use of distillationequipment can significantly reduce demands on fired heaters and reboilers. Theimportance of distillation is illustrated by the fact that in the U.K., distillation-relatedenergy consumption accounts for 13% of the total energy use by process industries.1 At

1 Chemical Engineering; July 1997; page 72.

Page 73: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2959

the same time, the expanding availability of process simulation computer packages hasprovided energy management teams with vastly increased analytical capability.

Changes to the operation of distillation columns must be undertaken so as to maintainproduct quality. There will usually be some deterioration in quality but in many casesthere is sufficient “over-purity” or “giveaway” that the product still meets specification.The following points should be considered when evaluating distillation columns.

The amount of overhead reflux. A considerable amount of heat is usually removed inthe top reflux exchangers. However, this is the coldest part of the column so there isless opportunity for transferring the heat to another product. By removing heat/moreheat using a pumparound stream located further down in the column, a hotter source ofheat is obtained and there is less heat removed in the overhead condenser.Pumparound streams are frequently installed on towers where there is no reboiler. Acommon example is a crude distillation tower.

The heat balance around the column dictates that, for constant product draw rates, lessheat put into the column means less heat has to be removed. For distillation columnsequipped with reboilers, adjustment of the reflux directly affects reboiler firing. (Note,even for reboilers that use a heat medium, the same principle applies because at somepoint in the process, a fired unit provides the heat to the circulating heat medium.)

Operating a propane-propylene splitter at a reflux ratio of 15.6 produces a propylenestream of 99.9% purity. Reducing the reflux ratio to 13.5 causes the propylene purity tofall to 99.7 mol %. However, the reboiler duty is reduced from 3,166 BTU/lb of propyleneto 2,740 BTU/lb of propylene. Fuel savings total 13.5%.1 Whether the operators couldcontrol the tower closely enough to achieve these savings is another point. Theimportant point of this example is the potential impact on energy use that even smalloperational changes can cause. Reflux can only be reduced as long as product qualityis acceptable.

Amount of overflash. This is another variation of reflux. When overflash rates (the liquidflowing down from the tray above the flash zone) are excessive, the flow can be reducedby lowering the feed or reboiler temperature or by raising the hydrocarbon partialpressure in the column. Note that pumparound streams affect the internal reflux flowsabove the pumparound trays. Overflash is unaffected by the existence ofpumparounds.2

Product splits. This is a measure of product purity. For light hydrocarbons, acomponent balance can determine whether the split between the light and heavy keycomponents is appropriate. For example, on a deethanizer in a gas plant the light keycomponent is ethane and the heavy key component is propane. There are usuallyspecifications on the amount of ethane allowed in propane. Ensure that thisspecification is being met, but not at the expense of distilling propane into the overheadgas. This not only consumes energy but downgrades product value.

The split between petroleum fractions is analyzed by comparing the distillation curves,especially the overlap of the tail end of the heavier product with the front end of the next

1 Chemical Engineering; December 1992; page 94.2 BP Research Project 139; Crude Oil Distillation, Principles and Practice; June 16, 1967; page 6.

Page 74: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 60

(lighter) product. The greater the fractionation, the smaller the overlap. However,allowable product quality should determine the size of the overlap.

The use of stripping steam. Steam is often injected into columns in order to reduce thehydrocarbon partial pressure. It is also used to strip out light material that may causeflash point problems.

As mentioned above, varying the hydrocarbon partial pressure of a distillation column –by using steam injection, for example – will change the overflash flow. Stripping steamrates should be set at the minimum required to meet specification. If stripping steamrates are too high, performance may actually deteriorate because of flow distributionproblems in the sidedraw strippers, causing channeling and flooding.

The points noted above are effective means of achieving energy savings at little or nocapital costs. There are also techniques for reducing energy use that involvefundamental changes to the distillation process itself. A common example of this is theinstallation of preflash columns. By only distilling one product overhead (sometimes twoproducts can be removed in a preflash column), less energy is required than in aconventional column with an overhead product and several product sidedraws.

As an example, a preflash column/crude distillation column arrangement was simulated(by computer) at 35,000 bpd of crude. By installing an 1,820 bpd sidedraw from thepreflash tower, overall energy requirements would have been reduced by 5.5 MMBTU/hr. This was equivalent to 3.2% of the total fuel input to the unit.1 Studies havebeen undertaken in Europe regarding the installation of a series of preflash towers oncrude units similar to the sequential use in gas plants of demethanizers, deethanizers,depropanizers and debutanizers.

Distillation columns/stabilizers are designed for specific operating pressures. Whencooling water and even more importantly, aerial coolers, are used for condensing theoverhead product, the design pressure is set by the maximum temperature reached bythe cooling medium. If the cooling water or ambient air becomes cooler the condenseroutlet temperature drops and the column pressure decreases. In older units, to take intoaccount situations where the water/air is cooler, the columns frequently have a bypassaround the condensers so that the design pressure can be held constant. Adherence tothis operating philosophy results in a lost opportunity to reduce energy costs.

Generally, it is desirable to operate at as low a pressure as possible in order to maximizethe relative volatility between the key components of the separation. The limit to whichthe pressure can be reduced is reached when a more expensive cooling medium isrequired.2

Consideration should be given to reducing the operating pressure of distillation columns.Those columns that utilize cooling water or air in the overhead condensers are potentialcandidates, especially if significant seasonal and/or diurnal changes in air/watertemperatures are experienced. As the temperature of the water/air changes, thepressure also varies, or “floats”.

1 BP Canada; Memo by N.A. Franklin; Preflash Sidedraw; February 28, 1978.2 GPSA Engineering Data Book, 10th Edition; page 19-4.

Page 75: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2961

As an example of this principle, floating pressure operation was instituted on a lightstraight run gasoline stabilizer in a crude unit in a Canadian refinery. The pressure wasreduced from the design of 145 psig to an average of 115 psig. Reboiler duty wasreduced by approximately 8.5%.1

As a second example, Figure 6.1 shows the effect of pressure on the reboiler duty of apreflash tower handling 34,400 bpd of crude oil.

Figure 6.1: EFFECT OF PRESSURE ON REBOILER DUTY2

0

5

10

15

20

25

15 20 25 30 35 40

Overhead Pressure, psia

Reb

oile

r D

uty

, MM

BT

U/h

ou

r ab

sorb

ed

Constant Reflux RatioNo Sidedraw Product

Investigate the feasibility of improving the efficiency of distillation column internals.Distillation column calculations deal with theoretical stages. The ratio of the number oftheoretical stages to the number of actual trays is the tray efficiency. As indicated inTable 6.1, there is a wide variation in tray efficiencies according to the service.Moreover, tray efficiency varies through the distillation column.

1 BP Canada, Memo by N.A. Franklin; Variable Operating Pressure for 21-C-6; December 7, 1976.2 BP Canada, Memo by N.J. Little; Sidedraw from No. 1 CDU Preflash Column; March 8, 1977.

Page 76: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 62

Table 6.1: TYPICAL TRAY EFFICIENCIES1

Demethanizer 45-60%

Deethanizer 50-70

Depropanizer 80-90

Debutanizer 85-95

Butane Splitter 90-110

Condensate Stabilizer 40-60

Crude Distillation Column* 52

Top 79

Middle 32

Bottom 67

Crude Preflash Column** 63

* BP Canada, memo by N.A. Franklin, 11-C-1, Primary Column Simulation,April 17, 1978** BP Canada, memo by N.J. Little, Sidedraw from CDU Preflash Column,March 8, 1977

The efficiency of distillation columns can be improved using packing. This can consist ofrandom packing – such as Pall rings, Raschig rings and Berl saddles – or structuredpacking – such as knitted-type mesh and corrugated plates.

Packing has several advantages:2

♦ Lower pressure drop per theoretical stage. For this reason packing (especiallystructured packing) is being used more and more in vacuum columns3

♦ High liquid loading. This is especially so for structured packing

♦ Corrosion protection. The packing can be made from ceramics and plastics,whereas trays may have to be fabricated from expensive alloys. Note that oneEuropean Company is now manufacturing glass-polytetrafluoroethylene bubble-captrays.4

1 GPSA Engineering Data Book, 10th Edition; page 19-16.2 GPSA Engineering Data Book, 10th Edition; page 10-17.3 Chemical Engineering; November 1997; page 37.4 Chemical Engineering; November 1997; page 37.

Page 77: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2963

Packing also has several disadvantages:1

♦ Limited turndown. Packing is limited to about 50% turndown, while trays can operateat 10-15% of full load.

♦ Liquid distribution problems. Channeling is much more prevalent in packed columnsunless there is proper distribution of the liquid at the top of the column and periodicredistribution throughout the column.

♦ Plugging. Dirt and other substances can more easily plug packing.

♦ Higher cost. Unless expensive alloys are required, trayed columns generally costless than packed columns.2

♦ Rigidity of design. It is very important to match actual performance with design whenconsidering packed columns. In other words, trayed columns are better suited tohandle deviations from design.3

Early distillation designs generally favored tray internals. During the energy crises of the1970’s and 1980’s, the energy savings achieved by packing resulted in the rapid growthof that technology. Lately, new tray designs are being brought into the market that canhandle greater liquid and vapor flows.4

In summary, determine the efficiency of the distillation columns at the facility. Whenconsidering means of improving efficiency, look at the question of trays versus packing.Also consult the vendors of both types of internals regarding the latest developments.

In closing this section, it is important to note that a considerable amount of work is beingundertaken in the area of distillation theory. It is more proper to say that many of thesetechnologies have been known for a long time but concerns about operation and controlof these units resulted in little acceptance by industry in the past. These concernsappear to be diminishing. Among the technologies now under development are thefollowing:

Fully thermally coupled columns link two or more columns with vapor and liquid streams.However, there are no reboilers or condensers between the columns. This arrangementreduces thermodynamic losses.

Divided-wall columns essentially have two fractionation columns mounted within a singleshell. This technology is a variant of the fully thermally coupled columns concept. Thisarrangement reduces the number of trays, column shells, reboilers and condensers. Inaddition energy consumption is lower.

Catalytic distillation combines reactions and distillation within a single column.

1 GPSA Engineering Data Book, 10th Edition; page 10-17.2 Chemical Engineering;November 1997; page 37.3 Ibid4 Ibid

Page 78: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 64

For more information on these subjects the reader is urged to read the following articles:

Chemical Engineering, Advanced Distillation Saves Energy and Capital, by F. Lestakand C. Collins, July 1997, pages 72-76.

Chemical Engineering, Catalytic Distillation Extends Its Reach, by K. Rock, G. Gildertand T. McGuirk, July 1997, pages 78-84.

Chemical Engineering, New Horizons in Distillation, by J. Humphrey and F. Seibert,December 1992, pages 86-98.

Other distillation technologies mentioned in the literature, such as the “spinning cone”and the “inverted columns” appear to have little potential use within the conventional oiland gas industry,1 at least in the near future.

6.2 Steam Systems

Energy management of steam systems is a complex subject. The reader is urged toread the ARPEL Guidelines for Energy Management of Steam Systems. This sectionwill briefly discuss the topic.

Unlike hydrocarbon systems, steam systems typically discharge, vent and release aconsiderable portion of the overall quantity of steam and condensate that is generated.Some of this lost steam/condensate is unavoidable but the energy management teamshould ensure that losses are minimized and that as much heat as possible is extractedfrom them prior to ultimate disposal. This will reduce the load on the boilers. Among thepotential energy management items are the following. (Efficient operation of the boilersis discussed in Chapters 5 and 7.)

Prepare steam balances for the various operating scenarios. These will highlight wheresteam is used, where it is wasted and where it is lost. On the basis of this information,priorities can be set.

Shut off steam going to units that are down or that no longer require steam. Steamtracing is frequently left in service long after it is no longer needed.

Ensure that steam traps are appropriate for the service and that they are in workingorder. This can be a serious source of steam loss. For example, studies have shownthat a steam trap that is leaking has been in that condition on average for six months.2

The average defective trap loses 50 pounds of steam per hour.3 While this seems aminimal amount, consider the number of traps in a facility. A survey of the Hüls Americachemical plant at Theodore, Alabama showed that 12.4% of its 887 traps were leaking.Steam losses totaled $336,000 US annually.4 At a Canadian heavy oil production facility

1 Chemical Engineering; Distillation Internal Matters; November 1997; page 37.2 Predictive Maintenance of Steam Traps: Combining Demand Side Management and Performance

Contracting; by F. Hooper, Jr. and R. Gillette; www.trapo.com/idea-2.htm.3 Back to Basics – Steam Traps 101; by D. Fischer;

www.powerspecialties.com/Armstrong_back_to_basics_Traps101.htm.4 Energy-Saving Steam Traps Earn Respect at Hüls; by R. Wily;

www.powerspecialties.com/EnergySavingSteamTraps.htm.

Page 79: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2965

over 20% of the steam traps were found to be leaking into the condensate return headerand causing excess steam venting in the deaerator.

Maximize the amount of condensate returned to the boiler feedwater. Loss ofcondensate not only represents energy waste but it means that water treatment costsmust increase because of the additional boiler feed water make-up.

Minimize the amount of boiler blowdown. A portion of the water in the boiler must bereleased in order to control the quantity of salts and minerals in the water. However,excessive blowdown wastes the energy input to raise the water from its inlet temperatureto that of the saturated steam, plus the associated pumping costs. Wherever possible,recover the heat in the blowdown before it is discarded.

Recover the heat in steam vented at the deaerator. Non-condensable gases, such asoxygen and carbon dioxide, must be vented from steam. This is typically done byinjecting low pressure steam into the deaerator and raising the temperature to near thatof saturated steam at atmospheric pressure or slightly higher. As a result steam isvented from the deaerator. At these pressures the latent heat of the steam is in theorder of 950-970 BTU/lb, whereas the enthalpy of the condensate is only 180-220BTU/lb (base = zero for water at 32°F). Condensing the vented steam from thedeaerator can provide heat to incoming boiler feedwater, or for other low-levelrequirements, such as a circulating heat medium.

Minimize steam sent to process. This must be done consistent with acceptable productquality or operating stability. Examples would include stripping steam in distillationcolumns and steam in FCCU regenerators.

Ensure steam-consuming equipment such as turbines and vacuum ejectors areoperating as efficiently as possible. Also ensure that the driven equipment (pumps andcompressors) is also operating both efficiently and effectively.

Determine the optimum steam pressure levels and steam quality (superheated,saturated, wet steam). This is a very complex exercise for the steam system mustaddress a wide variety of needs. A thorough knowledge of the steam system and theprocess requirements is needed.

Consider the use of condensing turbines. High pressure turbine exhaust steam meansthat the enthalpy drop per pound of steam is low so a large quantity of steam is needed.Condensing turbines greatly increase the enthalpy drop per pound of steam so thequantity of steam is reduced. To improve overall thermal efficiency, recover the heatfrom the condenser.

Evaluate the relative merits of operating electric motors instead of steam-driven turbines.Depending upon the location of the facility and the relative costs of electricity and fuel,motors may be more economical. Reliability of the power supply must be considered.

Ensure that the latent/sensible heat of steam is recovered for process use, rather thanremoved by cooling water or air.

Page 80: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 66

Identify sources of process heat for generating steam in waste heat boilers. Often thisheat is supplied by flue gas from heaters (see Section 7.3). Other sources include hotprocess streams such as vacuum residuum and FCCU regenerators.

6.3 Insulation

Proper installation and maintenance of insulation on piping and equipment is animportant component of an energy management program. Wet and damaged insulationshould be repaired as soon as possible.

Inadequate insulation can have severe repercussions on operations such as downholesteam injection in oil fields, due to the length of pipe involved. For estimates of the heatloss from piping, the reader is referred to Chapter 14 of the ARPEL guideline on EnergyManagement of Steam Systems. The graphs included therein are applicable to bothsteam and process lines.

Page 81: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2967

7.0 ENERGY RECOVERY TECHNIQUES

While not an infallible statement, it is generally true that older heaters have lower efficiency thannewer ones, unless they have been revamped. The techniques mentioned in Chapter 5 willimprove heater operation, but only within the confines of the equipment limitations. To attemptto operate a heater beyond these limits runs the risk of flame impingement, high metaltemperatures, improper draft profile in the heater and fans and pumps running at overloadconditions. Additional heater duty will require additional equipment.

It is stressed that prior to revamping a heater or boiler that a comprehensive analysis of theheater should be undertaken. Among the data that should be collected are the following:

♦ Process conditions (flows, temperatures, pressures and pressure drops)

♦ Fuel analyses

♦ Flue gas analyses

♦ Temperature and pressure profiles throughout the heater

♦ Tube temperatures, preferably including a survey using infrared thermography

♦ Complete operating data of heater peripherals such as waste heat recovery unitsand fans

♦ Known limitations

♦ Heater design data

♦ Heater inspection report

The above-mentioned data should be collected at the same time in order to obtain a “snapshot”of the heater.

There are three main methods of revamping heaters and boilers in order to improve thermalefficiency:

♦ Modify and/or expand the radiant and convection sections of the heater for greaterprocess heat

♦ Convert natural-draft to forced-draft operation

♦ Install waste heat recovery.

Each of these methods will be discussed in this chapter. Note that this review will not look atmethods of improving heater capacity that do not simultaneously improve heater efficiency.

Page 82: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 68

7.1 Modifying Radiant and Convection Sections

The performance of the radiant section is gauged by its heat flux and the velocity of theprocess fluid through the tubes. Typical values for refinery process heaters are providedin Table 7.1.

Table 7.1: HEAT FLUX RATES AND MASS VELOCITIES FOR REFINERY PROCESS HEATERS1

Service Radiant Flux Mass Velocity

BTU/hr/ft2 lbs/sec/ft2

Atmospheric crude heater 9,827–12,046 174–251

Reduced crude vacuum heater 7,925—9,985 61-102

Reboilers 9,510-12,046 143-251

Circulating oil heaters 7,925-10,936 348-451

Catalytic reformer charge reheat 7,449-12,046 45-72

Delayed coker 9,510-10,936 348-451

Visbreaker heating section 8,876-9,985

Visbreaker soaking section 6,023-6,974

Propane deasphalting 7,925-9,034

Hydrotreater charge 9,510-12,046 154-205

Hydrocracker charge 9,510-12,046 154-205

Steam superheater 9,827-14,899 31-76

If the radiant section is not meeting the design specifications, the investigators should:

♦ Ensure that the heater tubes are clean and that there are no flame impingement andhot spots, which could indicate coking on the inside of the tubes.

♦ Check the amount of air to the burners. As indicated in Figure 7.1, as the amount ofair is increased, the fraction of heat absorbed in the radiant section decreases.Remember that, for natural gas firing, stoichiometry requires approximately 10volumes of air per volume of gas. This means that roughly 17 pounds of air arerequired for every pound of fuel gas. At 15% excess air (19.5 lbs air/lb gas) and aflux rate of 10,000 BTU/hr/ft2 about 53.5% of the total allowable heat flux occurs inthe radiant section. Operating at 25% excess air, the radiant section absorbs only50.7% of the total heat.

1 H. Ghosh; Improve Your Fired Heaters; Chemical Engineering; March 1992; page 116.

Page 83: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2969

♦ Determine whether there is poor heat distribution within the firebox. When doing thisit is necessary to take into consideration the distance of the tubes from the flamesand the airflow in the firebox. The maximum radiant temperature is approximately 2-3 feet from the flame.1 At greater differences the temperature decreases. A long,narrow flame therefore, will have a relatively constant distance between it and thetubes in the radiant section. This results in a more even distribution of heat. Theheight of the flame should not exceed 50% of the height of the radiant section. As ageneral rule, the proper flame size is about 15 inches high for every million BTU/hr ofheat release. The diameter should be about 30 inches. To prevent flameimpingement, the flame edge should be 24-31 inches from the tubes.2

Natural gas burners produce long flames. This is due to the relatively poor mixing of thefuel and air. Low excess air rates at the burner or poor air distribution will exacerbatethis situation. The combustion is completed higher up in the firebox where there is moreoxygen. (This is the principle used in staged-air firing. See Section 9.3.)

Air leaks into the firebox will also change the heat distribution pattern not only by causingchanges to the airflow but also by decreasing the temperature of the gases. This lastpoint is important. Radiant heat transfer coefficients are functions of temperature to thefourth power. At the temperatures typically experienced exiting the radiant section(1500-1900°F) and the process temperatures (about 300-700°F), even a 10°F drop inthe firebox gas temperature decreases the radiant heat transfer coefficient by nearly 2%.Localized variations in firebox temperature will cause localized variations in heattransfer.

The firebox internal construction and air leaks may result in short-circuiting of the airflow,meaning that heat is taken out of the radiant section prematurely. It is very important toremember that heaters and burners are designed to meet the specifications of the client.Depending upon the desired process conditions, and the desired heat transfer rates,there could be long, narrow flames or short, wide flames, and a variety of gas flowregimes as the hot gases travel from the burners to the convection section. Moreover,present firing conditions can range from under-firing to over-firing; both of which couldexhibit radically different heat distribution patterns from design. It is recommended thatburner and heater manufacturers be consulted during analysis of the radiant section.3

The first two points mentioned above refer back to the discussion in Chapter 5.However, in order to properly assess the performance of the radiant section, it isnecessary to optimize conditions, within the limits of the available equipment. The thirdpoint lies outside the control of the operating staff and will require a capital outlay inorder to rectify it. It is recommended that heater manufacturers or specialists in this fieldbe consulted regarding modifications to the heater internals.

There are several options available regarding the convection section. An approachtemperature (flue gas outlet minus fluid inlet) of 90°F can be used.4 This can beachieved by:

1 Discussion with John Zink Canada.2 H. Ghosh, Improve Your Fired Heaters; Chemical Engineering; March 1992; page 121.3 Discussion with John Zink Canada.4 A. Garg; Revamp Fired Heaters to Increase Capacity; Hydrocarbon Processing; June 1998; page 72.

Page 84: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 70

♦ Adding tubes. Flux rates in the convection section are usually in the range of12,000-24,000 BTU/hr/ft2 of bare tube surface or 2,000-4,000 BTU/hr/ft2 of finnedsurface.1 (See the next paragraph.) Many heaters have extra space in theconvection section to allow the installation of two more rows of tubes. If this extraspace is not available, a common solution is to install the new tubes in the breechingleading to the stack.

♦ Install tubes with extended surfaces. Extended surfaces in these situations areusually helical fins, although studded tubes were commonly installed in the 1970’sand l980’s. Studded tubes can transfer 2-3 times more heat than bare tubes.Finned tubes typically transfer 8–11 times more heat than bare tubes. Fins offerthree advantages over studs: they transfer more heat; they cause much lesspressure drop and they are cheaper.

Figure 7.1: HEAT TRANSFER IN THE RADIANT SECTION OF DIRECT FIRED HEATER2

30

35

40

45

50

55

60

65

70

12 14 16 18 20 22 24 26 28 30 G, lbs air / lb fuel

Fra

ctio

n o

f T

ota

l Hea

t A

bso

rbed

in R

adia

nt

Sec

tio

n

6,000

9,000

12,000

15,000

Parameter is the Allowable Heat Flux to the Tubes, BTU/hr/ft2

Nominal Multiply Pipe Size "G" byinches 2 1.03 3 1.02 4 1.01 6 1.00 8 1.00 10 0.995

If considering the installation of tubes with extended surfaces, it is necessary to accountfor the different sizes of tubes and the tubesheets. If the heater has oil-firing it isnecessary to install soot-blowers between every four or five rows of tubes.

Note that the first two rows of tubes in the convection section (called shock tubes)receive radiant heat because they “see” the radiant flame, as well as convective heattransfer. In fact, the first shock row has the highest heat flux in the heater. The shockrows always have bare tubes. The third row of tubes receives radiant heat if it has longradius return bends and it should have bare tubes also. The first row of finned tubes isequipped with fewer, shorter and thicker fins in order to reduce the fin tip temperature.3

1 GPSA Engineering Data Book, 10th Edition; page 8-17.2 GPSA Engineering Data Book, 10th Edition; page 8-16.3 GPSA Engineering Data Book, 10th Edition; page 8-16.

Page 85: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2971

Installation of extra tubes will increase thermal efficiency by lowering the flue gas exittemperature. However, as a result, there is a greater flue gas pressure drop and there isless draft. Corrective measures are usually to extend the stack or to install an induceddraft fan at the top of the convection section. If the modifications to the heater result inmore fuel firing, the ability of the stack to handle the additional flue gas must bechecked. With the extra equipment installed, it will be necessary to ensure the structureand foundations can safely support the additional weight.

7.2 Conversion to Forced-Draft Air Supply

Natural draft burners are limited to low pressure drop across the burner - 0.3-0.6 inchesof water column (WC). Therefore, the air must be induced at low velocity. This leads torelatively poor air-fuel mixing. Excess air rates must be in the range of 15-20% for gas-firing and 30-40% for oil-firing.

Flame lengths are higher for natural-draft burners than for forced-draft burners. Gas-fired flames are about one foot per MMBTU/hr of firing. Oil-fired flames are about twofeet per MMBTU/hr.

Conversion of the air supply to forced-draft fans will increase heater efficiency byimproving air-fuel mixing so that there is better combustion with shorter flames that aremore stable. Forced-draft fans are a necessity if air preheating equipment is installed.

Forced-draft burners are discussed in Section 4.4. Suffice it to say, because of thebetter air-fuel mixing, use of a high pressure air burner will enable the heater to operateat considerably lower excess air rates (5-10% for gas-firing and 10-15% for oil-firing).See Section 3.4 for the effect of reducing excess air on heater efficiency. In addition, theflames are shorter and wider. This results in a more uniform heat distribution within thefirebox, as well as less danger of flame impingement.

It is important to remember that under normal circumstances, the conversion of burnersfrom natural draft to forced draft per se cannot be justified in terms of energy savings.However, by taking into account the ability to run the heater harder or at higher capacity,greater safety due to a lower risk of flame impingement and the ability to install airpreheating facilities, the conversion often becomes viable.1

A project to convert burners to forced-draft air supply must consider the space requiredfor the fan and ductwork. Included in this consideration is the space beneath the heater.Ducting to the burners (under the heater) and deeper windboxes will be needed.2

7.2.1 Forced-Draft Air Supply and Low-NOx Burners

Flame lengths for burners using natural draft were mentioned in Section 7.2. If theburners were replaced with low-NOx burners (still under natural draft) the flame lengthswould be 50-100% longer. Converting to a forced-draft air system will lead to shorter

1 A. Garg; Hydrocarbon Processing; June 1998; page 76.2 A. Garg; Hydrocarbon Processing; June 1998; page 78.

Page 86: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 72

flames. However, low-NOx burners in a forced-draft system will still have long flamesand low excess air rates (i.e., below design levels) will lead to operating problems.1

Conversion of burners to the low-NOx type may be necessary in order to meet air qualityregulations. To lessen the risk of flame impingement that may occur because of thelonger flame length, more burners can be installed. As discussed above, flame length isa function of the burner heat release. Therefore, more burners, for the same overallheater duty, will result in shorter flames.2

7.3 Waste Heat Recovery

Waste heat recovery is a common method if increasing the efficiency of a heater orboiler. The impact of such equipment is illustrated in Section 3.4. Five possible streamscan be heated using waste heat recovery:

♦ Combustion air

♦ Boiler feedwater

♦ Fuel gas

♦ Fuel oil

♦ Process fluid

Of these, the first two – combustion air and boiler feedwater – are the most common forprocess heaters and boilers, respectively. The temperature of flue gas leaving theconvection section will vary, frequently according to the age of the heater.Temperatures in the range of 350-1500°F can be found. The purpose of waste heatrecovery is to reduce the flue gas temperature, thereby increasing the thermal efficiencyof the heater. Every 35-40°F decrease in flue gas temperature is equivalent to a 1%increase in efficiency.

The minimum temperature to which a flue gas may be cooled is dependent upon theconcentration of acid gases in the stream. Of particular importance is the concentrationof SOx, which in turn influences the amount of SO3 in the flue gas. Therefore, theamount of sulfur in the fuel oil and/or H2S in the fuel/natural gas is a parameter that mustbe monitored closely.

Figure 7.2 shows the relationship between sulfur/H2S content and the acid gas dewpoint. It is drawn from two sources. As indicated in the drawing the literature values forfuel oil (at least) show a diversity. In the absence of fuel-specific dew point data, it issuggested that the highest dew point - gas or oil – for the reported sulfur/H2S content beused, in order to be conservative. To prevent corrosion due to localized cooling of theflue gas it is recommended that the metal temperature be kept at least 25°F higher than

1 Ibid, page 76.2 Ibid, page 78.

Page 87: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2973

the acid gas dew point.1 Manufacturers of waste heat recovery equipment may havetheir own design criteria.2

Figure 7.2: ACID GAS DEW POINT AS A FUNCTION OF FUEL SULFUR/H2S CONTENT3

150

170

190

210

230

250

270

290

310

330

0 1 2 3 4 5

H2S in Fuel Gas or Sulphur in Fuel Oil, %

Aci

d G

as D

ewp

oin

t, °

F

Maximum Flue Gas Dewpoint for GasExcess Air = 5-20% (Garg)

Fuel Oil(Garcia-Borras)

Maximum Flue Gas Dewpoint for OilExcess Air = 10-30% (Garg)

One method of avoiding dew point problems is to heat the cold fluid using a warmstream prior to it entering the preheater. This, of course, requires a warm stream closeby. Another option is to have the inlet of the cold fluid (A) at the bottom of the preheater(where the flue gas is the hottest). Once stream A has been warmed to a certain extentso as to avoid dew point problems, it is then routed to the top of the preheater (wherethe flue gas is coldest).

7.3.1 Combustion Air Preheat

Air preheaters consist of several types:

♦ Regenerative

♦ Recuperative

♦ Circulating liquid

The regenerative type is a technology adapted from boilers. A common example of thistechnology is the Ljungstrom heater. It consists of a slow-moving rotor packed withmetal plates or wires. Partitions, and radial/circumferential seals separate the hot fluegas and the combustion air. Flue gas temperatures up to 1500°F (1800°F, if specialalloys are used) and gas velocities of 500 ft/minute can be handled. The effectivenessof these units can reach 85-90%.

1 A. Garg; Hydrocarbon Processing; June 1998; page 80.2 Maxim Heat Recovery Applications Manual; as an example.3 A. Garg; Hydrocarbon Processing; June 1998; page 78 and Garcia-Borras, page 32

Page 88: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 74

The main drawbacks to this equipment are there are moving parts, leakage and powerconsumed turning the rotor. Leakage is due to:

♦ Entrainment in the rotor passages (this can be reduced by installing a blow-outsection between the hot and cold streams)

♦ The gap between the rotor shell and its housing

♦ Gaps in the radial seals.1

New types of regenerative units have improved seals. Large preheaters haveautomatically-activated deflecting sector plates at the hot end.2

Recuperative waste heat recovery units are gas/gas static exchangers. They are oftenequipped with fins. Some have fins on the inside and outside of the tubes and some onthe outside only. The type and location of the fins are determined by the possibility ofacid gas dew point problems.

A technology that is relatively new to the oil and gas industry, although it has been usedin petrochemical plants, is the heat pipe. This is usually a chrome-coated sealed coppertube filled with a volatile liquid. Usually it is ammonia but low boiling point hydrocarbonshave been used. A number of these tubes are mounted in a tubesheet that separatesthe hot and cold fluids. The tubes are mounted at a non-horizontal angle, with the hotend mounted at the lower elevation. There are no moving parts.

The hot flue gas passing over the tube vaporizes the ammonia, which flows to the top(cold end) of the tube. The cold air condenses the amount, thereby extracting the heatof vaporization. The liquid ammonia then flows by gravity down the tube and the cyclerepeats itself.

Projects to install air preheat facilities must take into account the space that will berequired for the ducting, the induced-draft fan on the flue gas outlet of the preheater, theforced-draft fan on the air inlet to the preheater, new burners and the preheater itself.The induced-draft fan can be eliminated by installing a circulating-liquid preheater. Theliquid is heated in the convection section and heats the air in a heat exchanger. SeeSection 7.3.4.

Recuperative air preheaters are larger and more expensive than regenerative units.However, they require less maintenance. See Figure 7.3 for a typical air preheatarrangement.

It is very important to remember that by preheating the air (or fuel), the adiabatic flametemperature increases. This means that the firebox and the tube walls will be hotter. Itwill be necessary to rerate the heater.3

1 Perry; pages 9-63, 9-64.2 A. Garg and H. Ghosh; Make Every BTU Count; Chemical Engineering; October 1990; page 218.3 A. Garg; Hydrocarbon Processing; June 1998; page 78.

Page 89: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2975

Figure 7.3: SCHEMATIC ARRANGEMENT OF AN AIR PREHEATER

Hot Flue Gas

Cold Flue Gas

Fired Heater

AirPreheater

ColdAir

Hot Air

Induced-Draft Fan

Forced-Draft Fan

Stack Damper

7.3.2 Boiler Feedwater Preheat

This is a very common option in steam plants. They are normally called economizers.For every 10°F rise in boiler feedwater temperature, the boiler duty is decreased byabout 1%. The tubes, usually arranged in a series-flow bundle are installed in the boilerbreeching. The tubes are generally finned. Temperature control is achieved byopening/closing a damper that is installed in a by-pass duct. Boiling in the economizermust be avoided.1

7.3.3 Fuel Preheat

Preheating of heater fuel (oil or gas) inputs energy into the firebox, similar to preheatingair. Therefore, the flame temperature will increase and the heater must be rerated.There is the additional safety precautions that must be taken since hot flue gas, whichcontains some air could come in contact with combustible material if the preheaterleaked.

7.3.4 Circulating-Liquid Heat Exchangers

Heat exchangers are one way of supplying heat to combustion air from the flue gas.They are also useful for supplying heat to circulating heat media and even to processand utility streams. (As with fuel preheaters, safety controls must be installed to preventdirect contact of flue gas and hydrocarbon streams. The use of a circulating mediumprovides a buffer between the two streams.)

Circulating-liquid exchangers are commonly used in place of air preheaters if the plantalready has a circulating heat medium or if space is limited. The option does not requirean induced draft fan or flue gas ducting.

1 Energy Management Handbook, 3rd Edition; page 215.

Page 90: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 76

Exchangers are being designed using transfer surfaces made of Pyrex. This allows theexchanger to cool the flue gas below its dew point. Plate exchangers can also cool fluegas to very low temperatures. Both sensible and latent heat is recovered from the watervapor in the flue gas. With proper design, the flue gas leaving the exchanger can still beslightly superheated so that there is no corrosion in the stack.1

7.3.5 Generation of Steam

Waste heat from flue gas is often used to generate steam, by installing tubes in theconvection section of the heater. The flue gas can be cooled to within 50°F of the boilerfeedwater temperature when generating low and medium pressure steam.2 The boilerfeedwater temperature must be hot enough to prevent condensation of the acid gas.

In many cases an induced-draft fan and flue gas ducting is required to install the wasteheat boiler. If using the waste heat to heat boiler feedwater, generally only minormodifications to the process heater convection section are needed.

Waste heat boilers are often installed on pyrolysis heaters and steam-reforming heaters.Steam is consumed in both processes.3

7.3.6 Waste Heat Recovery from Incinerators

The ability to recover heat from an incinerator will depend upon the size of theincinerator, the type of waste being burned and the need to have good plume rise anddownwind dispersion. Incinerators that completely destroy their wastes have moreoptions available to them. On the other hand, tail gas incinerators in sulfur recoveryunits destroy H2S but still produce SOx so the plant must retain the ability to emit thestack gases in accordance with the air quality regulations. Nevertheless, energymanagement opportunities may exist. The reader is referred to the following articles:

Chemical Engineering; Capture Heat From Air Pollution Control; by J. Straitz; October1993; pages 6-14.

Hydrocarbon Processing; Recover Heat from Waste Incineration; by V. Ganapathy;September 1995; pages 51-56.

7.4 Heat Sinks for Cyclic Operations

As discussed in Chapter 4, hot oil and molten salt heaters are used to provide heat forservices such as circulating heat media and heat regeneration gas for desiccant dryers.When these services are intermittent, there may be opportunities for saving energy.

Regeneration of molecular sieve beds and desiccant beds is a cyclic procedure involvingnormal operation, heating to regenerate the bed followed by cooling prior to switchingback to normal operation. Even with two or three beds, there will probably be timeswhen no bed is being heated. It is not practical to shut down the regenerator gas heater

1 A. Garg, H. Ghosh; Chemical Engineering; October 1990; page 220.2 A. Garg; Hydrocarbon Processing; June 1998; page 78.3 Ibid.

Page 91: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2977

for these relatively short periods of time. Therefore, the regeneration gas is heated andpassed over a heat-retentive sink such as lava rock. The stored heat is then used tostart the regeneration of the bed in the next cycle. Since the regeneration requires afixed amount of heat input, the overall firing on the heater can be reduced by the amountof stored heat transferred to the sieve/desiccant bed.

Energy use in such cyclic operations can be also reduced by extending the cycle lengthto the maximum possible, consistent with good operation. For instance, on desiccantbeds, the normal operation cycle should be extended until there is water breakthrough.In this case the savings are realized by the reduced number of times that the bed itself(i.e., the molecular/sieve/desiccant) and the bed shell must be heated to theregeneration temperature. The energy savings are relatively minor, compared with thereduced cost of replacing the desiccant or molecular sieve. There are also energysavings in circulating the regeneration gas because a lower flow rate is required.

Page 92: Optimizacion de Combustion ARPEL
Page 93: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2979

8.0 EMISSIONS FROM HEATERS AND BOILERS

Heaters and boilers in the oil and gas industry primarily use three fuels: natural gas, heavy fueloil and fuel gas. The term “fuel gas” refers to the off gases produced by many processes. Inmost circumstances in the upstream industry, the fuel gas is so similar to natural gas that itcould be considered to be natural gas. However, the internally-produced fuel gas in refineriescan be quite different. Often, it contains alkenes and much more hydrogen. Many heaters havedual-fuel capability.

The environmental impact of emissions resulting from combustion is of increasing concern. Asa result, initiatives such as energy management, acid rain reduction and ozone abatement arebeing developed in many jurisdictions in order to improve air quality. This chapter providesemission factors for the fuels commonly used in heaters and boilers. Chapters 9-11 look atemission control technologies for reducing the major air contaminants associated withcombustion. It must be stated that these control technologies will, at times, have an adverseeffect upon the energy efficiency of the facility. Thus the need for effective energy managementis even greater.

8.1 Emission Factors

The published emission factors are generally stated as functions of either the amount offuel (especially for residual fuels, in the form of weight/volume of oil) or heating value.The data presented here are a synopsis of the information provided in the ARPELGuidelines for Atmospheric Emissions Inventory Methodologies in the PetroleumIndustry. It is important to note that these factors are based upon typical operatingconditions and equipment. Wherever possible heater-specific emission factors or directcalculations/measurements should be taken.

Table 8.1 lists emission factors for heaters firing natural gas. Table 8.2 deals withemissions resulting from the firing of refinery fuel gas. Two scenarios are provided: lowhydrogen-content and high hydrogen-content fuels. The implied compositions are basedupon a two-train refinery that had separate fuel gas headers. Note that refinery fuel gascan vary significantly in hydrogen content if there is a catalytic reforming unit, due to theeffects of catalyst deactivation.

Table 8.1: EMISSION FACTORS FOR FIRED HEATERS (NATURAL GAS), lbs/MMBTU1

Heater Duty,

MM BTU/hr

<9.9 9.9-99 >99

CO2 116.93 116.93 116.93

CO 0.035

CH4 0.00267 0.00298 0.00298

NMHC 0.00279

NOx – no air preheat 0.0996 0.1393 0.5473

1 CAPP Voluntary Challenge for CO2, CH4, NOx; Canadian Ministry of Energy, Mines and Resources for

NMHC, CO, N2O, SO2, US-EPA 4/93, Table 1.3-10, for particulates.

Page 94: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 80

with air preheat 0.1661

N2O 0.00008

Particulates 0.1000

SO2 0.00047

NMHC Non-methane hydrocarbons

Table 8.2: EMISSION FACTORS FOR FIRED HEATERS (REFINERY GAS), lbs/MMBTU1

Heater Duty,

MM BTU/hr

<9.9 9.9-99 >99

Low Hydrogen Content Gas

CO2119.62

119.64 119.64

CO 0.035

CH4 0.00061 0.00068 0.00068

NMHC 0.03311

NOx – no air preheat 0.0996 0.1393 0.5473

with air preheat0.1661

N2O 0.00008

Particulates 0.1000

SO2 0.00037

High Hydrogen Content

CO2 69.73 69.74 69.74

CO 0.035

CH4 0.00045 0.00050 0.00050

NMHC 0.01863

NOx – no air preheat 0.0996 0.1393 0.5473

with air preheat0.1661

N2O 0.00008

Particulates0.1000

1 Adapted from CAPP Voluntary Challenge for CO2, CH4, NOx; Canadian Ministry of Energy, Mines and

Resources for NMHC, CO N2O, SO2; US-EPA 4/93, Table 1.3-10 for particulates.

Page 95: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2981

SOx Trace

NMHC Non-methane hydrocarbons

Note that NOx emissions will be a function of the adiabatic flame temperature, which is inturn a function of the fuel composition and the oxygen content of the flue gas. SeeSection 1.3. Wherever possible actual values should be calculated. Similarly, emissionsof carbon monoxide can be estimated using the data presented in Section 9.2.

Emission factors for oil firing will depend upon the type of oil (i.e., No. 4, No. 5, or mostoften, No. 6 Oil). See Table 8.3. Note that some of the factors are expressed in termsof lbs/barrel of fuel oil and others are expressed in terms of lbs/MM BTU.

The category “heavy metals” consists of at least eleven metals, such as antimony,arsenic, cadmium, chromium, cobalt, lead, manganese, nickel and selenium.Approximately 75% of the emissions of heavy metals are nickel. For further informationon this subject and others pertaining to emissions from heaters and boilers, see Section6.1 of the ARPEL Guidelines for Atmospheric Emissions Inventory Methodologies in thePetroleum Industry.

Emission factors for distillate fuels have been included. Normally, distillate fuel wouldnot be fired alone. However, it is used as a cutterstock. The values can be used whenblending No. 6 oil with cutterstock to make improved fuel oils. See Section 11.1.

Table 8.3: EMISSION FACTORS FOR AIR CONTAMINANTS FROM UNCONTROLLED RESIDUALOIL COMBUSTION1

Units No. 2 Oil No. 5 Oil No. 4 Oil No. 6 Oil

Utility Boilers SO2 Lbs/bbl 6.66S 6.31S 6.66S

SO3 Lbs/bbl 0.2418S 0.2418S 0.2418S

NOx* Lbs/bbl 2.804(1.753)**

2.804(1.753)**

2.804(1.753)**

CO Lbs/bbl 0.210 0.210 0.210

PM Lbs/bbl 0.421 0.294 0.3926S +0.130

CH4 Lbs/bbl 0.0119 0.0119 0.0119

NMHC Lbs/bbl 0.0319 0.0319 0.0319

Industrial Boilers

SO2 Lbs/bbl 5.96S 6.66S 6.31S 6.66S

SO3 Lbs/bbl 0.0841S 0.0841S 0.0841S 0.0841S

NOx* Lbs/bbl 0.841 2.313 0.841 2.313

1 US-EPA 4/93, Chapter 1.3.

Page 96: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 82

CO Lbs/bbl 0.210 0.210 0.210 0.210

PM Lbs/bbl 0.084 0.421 0.294 0.3926S +0.130

CH4 Lbs/bbl 0.0021 0.0421 0.0021 0.0421

NMHC Lbs/bbl 0.0084 0.0119 0.0084 0.0119

All Boilers

N2O Lbs/bbl 0.00456 0.00456

POM Lbs/MMBTU 0.000022 0.0000074 -0.0000084

FormaldehydeLbs/MMBTU 0.000233 -

0.0004050.000161 -0.000405

Heavy

Metals

Lbs/MMBTU 0.00011 -0.00013

0.00109 -0.00329

PM particulate matterNMHC non-methane hydrocarbonsPOM polycyclic organic matterS wt % S in fuel* If nitrogen content is known, use 0.864 + 4.39 * N, where N is the weight % of nitrogen in

the fuel. For vertically-fired utility boilers at full load and 15+% excess air, use 4.416.** Bracketed numbers are for tangentially fired utility boilers.

8.2 Effect of Heater Size on Emissions

Note in Tables 8.1 and 8.2 that heater size affects the emissions of methane and NOx.US-EPA Document 4/93, Chapter 1.3, Fuel Oil Combustion, lists emission factors for“utility boilers: and “industrial boilers”. The document defines neither term. It isassumed that a “utility boiler” is a large fired unit in a steam plant or a public electricalgeneration facility. The term “industrial boiler” is assumed to be a smaller boiler in asteam plant, and also applicable to process heaters. Despite the confusion regardingthe terminology, it is clear that the size of the boiler is a parameter affecting the quantityof emissions.

In their study of NOx emissions from heaters in steam reforming units, Kunz et al foundthat the US-EPA AP-42 factors overestimated emissions of NOx and CO.1 In earlierpublications on this subject (NPRA Paper AM-92-56; Control NOx from Gas-FiredHydrogen Reformer Furnaces; March 1992 and Hydrocarbon Processing; Control NOx

from Furnaces; August 1992) they attribute this over prediction to the fact that the AP-42factors were derived from studies done on utility boilers and are thus not valid for typical(and smaller) boilers and process heaters.

In summary, no definitive and comprehensive relationship between heater/boiler sizeand emissions was found in the literature. However, studies have found size to have a

1 Kunz et al; Hydrocarbon Processing; November 1996; page 76.

Page 97: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2983

bearing on the quantity of several categories of emissions. Whenever possible, heater-specific emission factors should be prepared.

8.3 Effect of Controls on Emissions

The emission factors listed in Section 8.1 are for uncontrolled emissions. That is, noequipment or operating procedures are in place to reduce emissions. The mostcommon control equipment involves measures to reduce emissions of NOx, SOx andparticulates. These are listed in Tables 8.4, 8.5 and 8.6, respectively. Descriptions oftechnologies and procedures are provided in Chapters 9 and 10.

Table 8.4: EFFECTIVENESS OF NOX CONTROL MEASURES FOR RESIDUAL FUEL COMBUSTION1

Technique Process Reduction Efficiency, %

CombustionModifications

Low excess air Reduce excess 02 to 2.5% 0-28

Staged combustion Fuel-rich burners, secondarycombustion air ports

20-50

Burners-Out-Service Some burners fuel rich, someburners air only

10-30

Flue gas recirculation 25-30% of the fuel gas recycled toburners

15-30

Flue gas recirculation plusstaged combustion

25-53

Load reduction Reduce air and fuel to all burners 33 to an increase of 25

Low-NOx burners New burner design 20-50

Reduced air preheat Bypass air preheater 5-16

Post-CombustionModifications

Ammonia injection NH3 injected into flue gas 40-70

Urea injection Urea injected into furnace 30-60

Thermal DeNOx NH3 injected into furnace 30-60

Air heater baskets Baskets of catalyst to promotereaction of ammonia with NOx

40-65

Selective catalyticreduction (SCR)

Catalyst in flue gas stream 90

Duct (SCR) Small version of SCR in existing duct 30

1 US-EPA 4/93, Table 1.3-12.

Page 98: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 84

Activated carbon on SCR Carbon catalyst downstream of airpreheater

Not available

Table 8.5: POST-COMBUSTION SO2 CONTROLS FOR RESIDUAL OIL COMBUSTION1

Technology and Chemical Agent Reduction Efficiency %

Wet scrubber

Lime/limestone 80-95+

Sodium carbonate 80-98

Magnesium oxide/hydroxide 80-95

Dual alkali 90-96

Spray Drying

Calcium hydroxide slurry 70-90

Furnace injection

Dry calcium carbonate hydrate 25-50

Duct injection

Dry solvent injection 25-50+

Table 8.6: REMOVAL EFFICIENCY OF TECHNOLOGIES FOR CONTROLLING EMISSIONS OFPARTICULATES2

Technology Removal Efficiency, %

Electrostatic precipitator 99.2

Scrubber 94.0

Multiple cyclones 80.0

1 US-EPA 4/93, Table 1.3-14.2US-EPA 4/93, Tables 1.3-5 and 1.3-6.

Page 99: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2985

9.0 CONTROLLING EMISSIONS – COMBUSTION CONTROLS

9.1 Fuel Switching

The emission factors listed in Tables 8.1, 8.2 and 8.3 indicate the impact that fuel typehas on emissions. As an example, consider Scenario 5 in Section 3.4. For that case, 40MM BTU/hr of process heat input is achieved using primarily direct firing, with a minoramount of waste heat recovery. The flue gas was also used to preheat the air and fuels.In Scenario 6, the fuel is switched to 100% gas firing and in Scenario 7, to 100% oilfiring. (Fuel qualities are listed in Section 3.4).

Table 9.1 lists the emissions for all three cases. The use of gas, because of its highercontent of hydrogen, results in considerably less emission of all the listed aircontaminants. The efficiency of combustion using gas is slightly lower because fullutilization of the heat content of the water vapor formed during combustion cannot berealized. However, no atomizing steam is required, thereby reducing fuel firing in thesteam plant. Note that the values in Table 9.1 are for the process heater only. Otherbenefits of switching from heavy fuel oil to gas are that no corrosion inhibitors, heating offuel oil lines and tanks and viscosity reducing additives are required. The net effect isthat switching to gas reduces fuel firing.

Emissions of NOx have been estimated using the factors for oil provided in Chapter 8and the factors for gas provided in Section 1.3. See Section 9.3 for further details.Emissions of carbon monoxide are based upon Figure 9.1. Carbon dioxide and SOx

emissions are calculated using carbon and sulfur balances around fuel and flue gasstreams.

Table 9.1 is prepared on the presumption that the heater could operate using either gasor oil or both. If the heat release per burner is 250,000 BTU/hr or greater, a dual-fuelburner is usually possible. Oil-derived flames have different heat release patternsbecause their flames are more luminous than those that are gas-derived. Oil flamesrequire more combustion space because additional time is required to ensure fuelatomization and vaporization. Most conventional general-purpose heaters are designedconservatively, (i.e., large firebox) so there is usually not a problem with excessive flameradiation.1

Another form of fuel switching involves the conversion of No. 6 Oil to No. 5 Oil or evenNo. 4 Oil. However, there are product yield implications with this choice. The reader isreferred to Chapter 3 of the ARPEL Guidelines on the Impact of Fuel Switching onRefinery Operations and Atmospheric Emissions.

1 North American Combustion Handbook; page 42.

Page 100: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 86

Table 9.1: EFFECT OF FUEL SWITCHING ON HEATER EMISSIONS, lbs/hour

Scenario 5 Scenario 6 Scenario 750% Gas, 50% Oil 100% Gas 100% Oil

CO26911

5917 7906

N2O 0.017 0.004 0.033

CH4 0.220 0.130 0.311

CO 1.819 1.380 2.277

SO2 37.205 0.679 73.707

SO3 0.470 0.009 0.930

NOx 24.235 16.440 32.315

NMHC 0.106 0.121 0.088

Particulates 4.612 4.376 4.846

Heavy Metals 46.756 0 93.476

NMHC Non-methane hydrocarbons

9.2 Excess Air vs. CO, NOx, Efficiency

The importance of maintaining good control of the amount of excess air has beendiscussed in Chapters 3 and 5. There is also an important reason concerning emissionsof NOx and carbon monoxide.

The decrease in heater efficiency brought about by the heat requirements of the excessair creates an incentive to operate with as low an excess air rate as possible. At thesame time, there is not always good air-fuel mixing at low excess air rates. This causesincomplete combustion and there is a steady accelerating increase in emissions ofcarbon monoxide with falling excess air rates. See Figure 9.1 for the relationshipbetween oxygen in the flue gas and emissions of CO. Fuel type is a factor.Instrumentation exists to control CO emissions in the 150-250 ppm range.

Page 101: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2987

Figure 9.1: CARBON MONOXIDE EMISSIONS FROM HEATERS1

0

200

400

600

800

1000

-1 0 1 2 3 4 5 6 7Oxygen Content of Flue Gas, vol %

CO

Em

issi

on

s, p

pm

Coal

Oil

Gas

CO Control Range

As stated in Section 1.3, the formation of NOx in gas-fired heaters and boilers is afunction of the oxygen content of the flue gas and the adiabatic flame temperature(AFT). At very low concentrations of oxygen in the flue gas (i.e., less than 1-2%) there isinsufficient oxygen to drive the formation of NOx forward. At concentrations of 4% andhigher, there is enough excess air to materially affect the adiabatic flame temperature.This “quenching” of the AFT outweighs the effect of the increased availability of oxygen.However, at middle concentrations of oxygen (2-4%), there are sufficient oxygen andsufficient adiabatic flame temperature present such that the generation of NOx is usuallymaximized in that range of oxygen content. See Figure 9.2.

Figure 9.2 illustrates the relationship between NOx, CO and heater efficiency at low ratesof excess air. Natural gas fuel is illustrated. For fuel oil, the heater efficiency line will beessentially unchanged. The lines for NOx and CO will be moved upward (i.e., greateremissions for the same oxygen content in the flue gas. Note also that the conditions thatgenerate the highest concentration of NOx do not generate the highest emission rate.This is because the amount of excess air is increasing faster than the rate that the NOx

concentration is decreasing. See Figure 9.3.

1 Garcia-Borras; page 31.

Page 102: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 88

Figure 9.2: EFFECT OF OXYGEN ON HEATER EMISSIONS

0

100

200

300

400

500

600

0 1 2 3 4 5 6 7

Oxygen in Flue Gas, volume % (dry basis)

Em

issi

on

s, p

pm

v (

dry

bas

is)

89.5

90.0

90.5

91.0

91.5

92.0

92.5

Hea

ter

Eff

icie

ncy

, % (

LH

V B

asis

)

Heater Efficiency

NOx

CO

Fuel is Natural Gas

9.3 NOx Abatement

Generally speaking, the lighter the fuel the less the NOx emissions, all other thing beingequal. An exception to this, however, are fuels rich in hydrogen. High concentrations ofhydrogen result in higher adiabatic flame temperatures. Abatement technologies usuallyinvolve combustion control - at the burner or post-combustion control - in the flue gas.

There are a great number of technologies available for reducing NOx. Some of thesehave been listed in Table 8.4. The effectiveness of these options will depend uponwhether they address the question of thermal NOx, fuel NOx or both.

Figure 9.3: NOx EMISSIONS FROM HEATERS

0

40

80

120

160

200

240

280

320

360

400

0 1 2 3 4 5 6 7

Oxygen in Flue Gas, volume % (dry basis)

Em

issi

on

s, p

pm

v (

dry

bas

is)

0

2

4

6

8

10

12

14

16

18

20

Em

issi

on

s, lb

/ho

ur

ppm v

lbs/hour

Fuel is Natural Gas

Page 103: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2989

9.3.1 Fuel NOx

There are two types of NOx – fuel NO x and thermal NO x. The first is the result ofcombustion of chemically-based nitrogen compounds in the fuel and the second is theresult of thermal fixation of nitrogen with the oxygen in the combustion air. Gaseousfuels contain very little nitrogen compounds, so fuel NOx is essentially a problemencountered only when burning fuel oils. Factors for NO x emissions are provided inSection 8.1. The source document for these factors (US-EPA 4/93, Chapter 1.3)estimates that 20-90% of the nitrogen in residual fuel oil is converted to NOx. This meansthat more than 50% (and in another part of the document, it is stated 60-80%) of the NOx

emissions from residual oil firing is fuel NOx. US-EPA 4/93, Chapter 1.3 is not explicitbut it is assumed that the factors for the combustion of residual oil include both thermaland fuel NOx.

Fuel nitrogen conversion to NOx is highly dependent upon the fuel-to-air ratio in thecombustion zone and, in contrast to thermal NOx formation, is relatively insensitive tosmall changes in combustion zone temperature. Increased mixing of air with the fuelincreases combustion efficiency but it also increases fuel NOx. To reduce this problem,initial air injection to the burners is kept below the stoichiometric amount. This results ina reducing atmosphere at relatively high temperature and long residence time. Fuelnitrogen is then converted to N2, rather than NO.1

9.3.2 Thermal NOx

US-EPA Document 4/93, Chapter 1.3 states that the thermal NOx concentration is afunction of the peak temperature (adiabatic flame temperature), fuel nitrogenconcentration, oxygen concentration and the residence time at the peak temperature(page 1.3-2). Kunz et al (Hydrocarbon Processing, November 1996) prepared formulaefor steam-reformer heaters that showed NOx formation to be a function of the adiabaticflame temperature and the oxygen content of the flue gas. (Since their study involvedgaseous fuel, the fuel nitrogen content would be extremely low and fuel NOx would notbe a significant factor.) Their formulae are presented in Section 1.3 of this guideline.

Production of thermal NOx is increased by better air/fuel mixing. Reduction of thermalNOx focuses on reducing the level of oxygen in the primary flame zone, by reducing theadiabatic flame temperature and by reducing the residence time at high flametemperatures.

9.3.3 Flue Gas Recirculation

A portion of flue gas, preferably less than 600°F,2 is recycled back to the burner throat.This achieves two goals: it dilutes the combustion air with dilute material, which lowersthe adiabatic flame temperature, and it reduces the concentration of oxygen, whichinhibits NOx formation.

Figure 9.4 shows the theoretical effectiveness of this technology, if it could be used to itsmaximum extent. Unfortunately, as outlined in the following paragraph, operationalproblems limit the amount of flue gas recirculation. Regarding this graph, note that the

1 US-EPA 4/93, Chapter 1.3; page 1.3-7.2 A. Garg; Chemical Engineering; November 1992; page 124.

Page 104: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 90

x-axis is in terms of the volume of flue gas recirculation per volume of natural gas fired.Normally, the recirculation rate is expressed in terms of percentage of flue gas. As anillustration, roughly 10 volumes of air are required for every volume of gas forstoichiometry. With 15% excess air, the flue gas flow is therefore, 11.5 + 1 = 12.5 timesthe gas flow. A recirculation ratio of 2 on Figure 9.4 is therefore, equivalent torecirculating 2/12.5 = 16% of the flue gas.

Figure 9.4: EFFECT OF FLUE GAS RECIRCULATION ON NOX EMISSIONS1

0

10

20

30

40

50

60

70

0 1 2 3 4 5 6 7 8 9 10 11

Recirculation Ratio

NO

x E

mis

sio

ns,

pp

m Recirculation Ratio = Volume of Recirculated Flue Gas per Volume of Natural Gas

The amount of flue gas recirculation is usually limited to 15-25% of the flue gas on gas-fired units, although a rate of 45% is possible on gas-fired boilers.2 At high recirculationrates, flame instability and high emission rates of particulates occur, especially on oil-fired boilers. For this reason, recirculation rates are typically limited to 10-12% for oilfiring.

There are several drawbacks to flue gas recirculation:

♦ Although a very successful technology for gas-firing (estimates in the literature rangefrom 50-85%), it is much less effective for oil-firing (10-15%). This is because itaddresses the issue of thermal NOx, whereas the majority of NOx with oil-firing is fuelNOx.

♦ This option is usually installed on heaters/boilers with forced-draft air supply. Itworks best in units with few burners, such as vertical, cylindrical heaters.3

♦ Capital costs can be quite high, especially on retrofits. Adequate space for theadditional ducting and fans may be an issue of concern.

♦ Continuous oxygen and carbon monoxide analysis is recommended. If the heatingvalue of the fuel is highly variable, a flame safeguard system should be installed.

♦ Corrosion in the ducting due to particulates, may occur.1

1 NPRA Paper AM-92-56, Figure 9.2 D. Fusselman, D. Lipsher; Oil and Gas Journal; November 2, 1992; page 48.3 A. Garg, page 126.

Page 105: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2991

♦ Flue gas recirculation does not affect the overall efficiency of the heater if thetemperature of the flue gas leaving the convection zone is the same as that of therecirculated flue gas. However, due to the diluent effect of the recirculated flue gas,heat transfer shifts from the radiant section to the convection section of the heater.This may be a problem for heaters operating near their maximum capacity.

9.3.4 Staged-Burners and Combustion

There are three main variations of this technology: staged-air burners, staged-fuelburners and staged-air combustion. The air or fuel, depending upon the variation, isadded in two stages, referred to as primary and secondary. (Some literature sourcesrefer to the air added after primary combustion as tertiary air.) Figure 9.5 showsschematic drawings of staged-air and staged-fuel burners.

Staged-air and staged-fuel burners are often called Low-NOx burners. Also listed underthis category are low excess air burners and ultra-low NOx burners. These are discussedin Section 9.3.5.

Staged-air burners inject all the fuel at the nozzle but the quantity of primary air is keptbelow the stoichiometric amount. Typically, burners are designed to give a residencetime in the primary combustion zone of 35-100 milliseconds.

1 Fusselman, Lipsher, page 48.

Page 106: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 92

Figure 9.5: STAGED BURNERS1

1 John Zink Company; NOx Control in Fired Heaters; paper by R. Martin, W. Johnson; presented at 1984 Winter National Meeting of the American

Institute of Chemical Engineers; Tulsa, Oklahoma; March 11-14, 1984.

Page 107: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2993

Approximately 65-70% of the stoichiometric air is added at the primary zone.1 The fuelpartially burns and the nitrogen forms ammonia, hydrogen cyanide and nitrogen oxide,which are subsequently reduced to elemental nitrogen. Incomplete combustionproduces a lower adiabatic flame temperature, as does the backmixing of the productsof combustion. The secondary air is added through the refractory ports to complete thecombustion and to stabilize the flame.

NOx reductions of 20-50% have been reported in the literature.2 The technology isrelatively inexpensive and has been used in forced-draft heaters and with flue gasrecirculation installations. The principal drawback is that the flame tends to be longerand wider than in conventional burners. This may cause flame impingement andthermal stresses.

Staged-fuel burners operate with all of the air, but only a portion of the fuel beinginjected through the burner. There are localized high levels of excess air, resulting inlow thermal NOx formation. Among the products of incomplete combustion are hydrogenand carbon monoxide, which reduce part of the NOx formed in the primary stage ofcombustion.

Staged-fuel burners produce lower NOx levels than staged air burners particularly in gas-fired heaters. Reduction of NOx is 50-70%. The flame is about 1½ t imes longer than instandard burners,3 but shorter than the flame produced in a staged-air burner.

Staged-air combustion is similar to the staged-air burner, except that the secondary air isinjected into the firebox. This means that the burners operate in a fuel-rich region andthe firebox in a fuel-lean region. Three variations predominate: overfire, burners-out-of-service, biased firing.

In the overfire air option, air injection ports are installed above the top row of burners.Burners-out-of-service is a commonly-used technique in retrofit situations where thereare multiple burners. In this option, the top row of burners receive air only (i.e., no fuel).In biased firing, only selected burners receive air. This is commonly used if the burners-out-of-service option results in firing limitations. These techniques (which in oneliterature source4 are all referred to as burners-out-of-service) are not practical forheaters with less than eight burners. Removing isolated single burners or a series ofburners in small heaters will cause flame pattern problems, reduced thermal capacityand reduced thermal efficiency.5

The feasibility of staged combustion depends upon the number of burners, the physicaldimensions of the firebox in order to install over-fire air ports and the residence time inthe firebox. The potential benefits, though, are significant: 60-70% reduction in NOx

when firing gas and 40-50% when firing oil or coal.6

1 John Zink Company; Burner Design Parameters for Flue Gas NOx Control; paper by R. Martin, no date;

page 15.2 A. Garg; page 128 states 20-35%; US-EPA 4/93, Table 1.2-12 states 20-50% for residual oil firing.3 A. Garg; page 128.4 NOx Control Techniques for the CPI; by D. Lambert, T. McGowan; Chemical Engineering; June 1996;

page 101.5 Ibid.6 R. McInnes, M. Van Wormer; Chemical Engineering; September 1990; page 133.

Page 108: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 94

9.3.5 Low NOx Burners

Low-NOx burners take advantage of the reaction kinetics of NOx formation and ensurethat high adiabatic flame temperatures and adequate oxygen content do not occursimultaneously. This is frequently achieved by staging either the injection of the fuel orthe addition of the combustion air. See Section 9.3.4 for more detail. Two othertechnologies are low-excess-air burners and ultra low-NOx burners.

Low-excess-air burners provide excellent mixing of the air and fuel. They can operateat flue gas oxygen concentrations of 0.5-1.5%. As seen in Figure 9.2, this is a regionwhere the oxygen concentration is limiting so NOx production is relatively low. Theseburners work best in single-burner heaters. With multiple burners there are difficultiesensuring proper and equal air distribution. They are mechanical or forced-draft burnersso they are more likely to be installed on boilers than on process heaters, which aremore often equipped with natural draft burners.

Savings of 0-28% have been reported. Also, because they operate at lower overallexcess air rates, fuel savings of 1-2% are realized.1

Ultra-low-NOx burners combine staged-fuel burners with flue gas recirculation. Theyrepresent a new technology. Production of NOx is very low. In general, installation oflow-NOx burners is an extremely popular and economical option, especially on newboilers. Retrofits are often more costly because of the changes needed toaccommodate the new burners.

Low-NOx burners operate in a reducing atmosphere in the firebox. This could lead toaccelerated slagging and corrosion unless the tube metallurgy is changed. Moreover,the flames are longer so flame impingement and thermal stresses may pose problems.In terms of emissions, the low excess air and incomplete combustion in parts of thefirebox could lead to increased generation of carbon monoxide, and smoke.

The impact of low-NOx burners upon NOx emissions, and even fuel efficiency, issubstantial. See this section and the previous one for examples. This is also showngraphically in Figure 9.6. The curves are from work by Kunz et al. (HydrocarbonProcessing, November 1996, pages 65-79). The exact type of low-NOx burner(s)involved in the study was not mentioned. Also, the curves are for gas-firing in a steam-methane reformer heater.

Figure 9.7 shows the effect of low-NOx burners on oil firing as a function of the weightfraction of nitrogen in the feed. Figure 9.8 shows the effect as a function of air preheattemperature. In both cases, high intensity burners are compared with low-NOx staged-air high intensity burners.

1 R. McInnes, M. Van Wormer, Chemical Engineering, September 1990, page 132.

Page 109: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2995

Figure 9.6: EFFECT OF LOW-NOX BURNERS ON NOX EMISSIONS (GAS FIRING)

0

50

100

150

200

250

300

350

0 1 2 3 4 5 6 7

Oxygen in Flue Gas, volume % (dry basis)

NO

x C

on

cen

trat

ion

, pp

m v

(d

ry b

asis

)Conventional Burners/Boilers

Conventional Burners in a Steam-Methane Reformer

Low-NOx Burners in a Steam-Methane Refomer

Figure 9.7: EFFECT OF NITROGEN CONTENT IN FUEL OIL ON NOX EMISSIONS USINGSTANDARD, HIGH INTENSITY / LOW NOX HIGH INTENSITY BURNERS1

0

100

200

300

400

500

600

700

800

900

1000

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0

Fuel Nitrogen, wt%

NO

x, p

pm

(d

ry b

asis

) co

rrec

ted

to

3%

O2

Conventional High Intensity Burner

Low-NOx High Intensity Burner with Staged Air

Air Temperature 90°F

1 Burner Design Parameters for Flue Gas NOx Control; Figure 18.

Page 110: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 96

Figure 9.8: EFFECT OF AIR PREHEAT ON NOX EMISSIONS (OIL FIRING) USING STANDARD HIGHINTENSITY AND LOW NOX HIGH INTENSITY BURNERS1

0

100

200

300

400

500

600

0 100 200 300 400 500 600

Air Preheat Temperature, °F

NO

x, p

pm

(d

ry b

asis

) co

rrec

ted

to

3%

O2

Conventional High Intensity BurnerOil Firing, 10% Excess Air

Low-NOx High Intensity BurnerOil Firing, 10% Excess Air

Low-NOx High Intensity BurnerGas Firing, 5% Excess Air

Nitrogen Content of Oil0.3 weight %

Conventional High Intensity BurnerGas Firing, 5% Excess Air

9.3.6 Diluent Injection

In this option, water or steam is injected into the combustion zone. Water is addedthrough nozzles in the windbox. This vaporizes it before it enters the combustionchamber. John Zink found that the most effective means of injecting steam was bymixing it with fuel gas. Steam is added directly to the combustion zone.

Significant reductions in NOx emissions can be achieved. One literature source statesthat a water-to-fuel ratio of 0.5 reduces thermal NOx formation by about 40%. Doublingthe water-to-fuel ratio reduces NOx by a further 20-30%.2 John Zink Companyconducted a number of tests and found that, by injecting a steam weight-to-fuel weightratio of 0.2-0.3, NOx reductions of 50% were achieved in a standard gas-fired burner.3

Figure 9.9 shows the effect of steam injection for a natural draft burner firing natural gas.Note that this graph shows a diminishing effect for steam injection ratios of greater than0.3. The Fusselman and Lipsher article quotes water-to-fuel ratios of 0.21 to 1:1 andsteam injection ratios of 1:1 to 2:1.4 The John Zink studies did find that the actualreduction will be a function of the burner design and the concentration of NOx in the fluegas prior to the steam injection.5 In view of the diversity of the results, it is suggestedthat a series of heater/burner-specific tests be conducted and the optimum injection ratebe determined empirically.

1 Burner Design Parameters for Flue Gas NOx Control; Figure 17.2 D. Fusselman and D. Lipsher; Oil and Gas Journal; November 2, 1992; page 46.3 Burner Design Parameters for Flue Gas NOx Control; by R. Martin; John Zink Company; page. 8.4 D. Fusselman and D. Lipsher, page 46.5 R. Martin, page 8.

Page 111: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2997

Figure 9.9: EFFECT OF STEAM INJECTION ON NOX EMISSIONS1

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0.00 0.05 0.10 0.15 0.20 0.25 0.30

Lbs .Steam/lb. of Fuel (CH4)

Rel

ativ

e N

Ox L

evel

s

For retrofits, injection of water or steam is usually a relatively cheap option. However, itis essential that the water be at least at boiler feedwater quality and that it be completelyvaporized before it enters the combustion zone. These conditions are necessary tominimize corrosion.

Newer boilers and heaters tend to operate at higher firing temperatures in order toimprove fuel efficiency. This results in greater NOx generation. At the same time, manyjurisdictions are reducing the allowable NOx emission limits. Therefore, in order toreduce emissions using this technology, higher injection rates are required. This has anadverse effect on fuel efficiency and CO emissions increase.2

As a variation of this technology, at least one company is conducting research into thecombustion of oil-water mixtures. Initial results are very promising.

1 R. Martin, Figure 5.2 D. Fusselman and D. Lipsher, page 48.

Page 112: Optimizacion de Combustion ARPEL
Page 113: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 2999

10.0 CONTROLLING EMISSIONS-POST-COMBUSTION CONTROLS

Considerable attention has been directed towards reducing emissions from heaters and boilersby treating the flue gas. The most common air contaminants of concern are NOx, SOx andparticulates. The first pollutant - NOx - can be treated either at the combustion stage or at post-combustion (flue gas) stage. SOx and particulates are commonly treated at the post-combustion stage. Improvement of fuel quality is another option and is discussed in the nextchapter.

Figure 10.1 shows schematic drawings of three “post-combustion” techniques for reducing NOx

emissions. Non-selective catalytic reduction and selective catalytic reduction are completelypost-combustion processes. Flue gas recirculation is partly post-combustion and partlycombustion.

10.1 NOx

Post-combustion technologies for reducing NOx emissions are usually either selectivenon-catalytic reduction (SNCR) or selective catalytic reduction (SCR). These areexpensive options to implement but they achieve very high NOx removal rates and areoften required where stringent emission regulations are in force.

Selective non-catalytic reduction uses ammonia or a urea reagent to combine with NOx

to form an intermediate ammonium salt, which decomposes to elemental nitrogen andwater. Injection of these chemicals is either into the firebox or the flue gas duct. Urea isbecoming more popular because it is safer and easier to handle. The urea decomposesfirst to carbon dioxide and ammonia, which then reacts with NOx to eventually formnitrogen and water.

Very close temperature control is required. If using ammonia injection, the flue gastemperature should be 1600-1750°F. Exxon’s Thermal DeNOx process has a statedoperating range of 1600-2200°F, when using ammonia alone. The addition of hydrogenplus ammonia extends the operating range down to 1300°F.1 If urea is used thetemperature range is much wider, 1000-1900°F.

If the temperature is too high, the ammonia will preferentially react with the oxygen in theflue gas and actually produce NOx. If the temperature is below the optimum range, theNOx conversion reactions slow down. This will lead to poorer NOx removal and aphenomenon called ammonia slip, in which unreacted ammonia is emitted in the fluegas. This in itself will lead to air quality problems.

NOx removal is typically 40-70%. It is an option well suited for retrofits. Capitalinvestment is low, compared with the catalytic version, and less space is needed. In astudy of SCR and SNCR for an oil-and-gas-fired boiler, the capital cost of the SNCRoption (Exxon’s Thermal DeNOx process) was only 20% of the cost of the SCR

1 Exxon Thermal DeNOx Process; brochure; Florham Park, New Jersey, USA; no date.

Page 114: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 100

scenario.1 The reagent can be injected directly into the boiler/heater. There are nohazardous waste concerns.

On the other hand, there are problems with SNCR if close operational control is notachieved. Ammonia slip has been kept as low as 10 ppm2 but if the proper ratio ofammonia to NOx is not maintained, slip can reach 50-100 ppm.3 The ammonia will reactwith a number of contaminants at low temperatures, such as sulfur and chlorine, to formcomplex salts. At flue gas temperatures below 600°F ammonia reacts with sulfurtrioxide and water vapor to form ammonium bisulphate and ammonium sulphate. Thesecan result in fouling of downstream waste heat recovery equipment and the bisulphatecan cause corrosion. Periodic water washing easily removes the sulphates. At flue gastemperatures below 250°F the ammonia reacts with hydrochloric acid to form ammoniumchloride. This may result in a visible plume.4

Selective catalytic reduction (SCR) uses the same principle as selective non-catalyticreduction (SNCR). Either anhydrous or aqueous ammonia can be used. However, thepresence of the catalyst reduces the required temperature range to 500-950°F. Thelimits are actually smaller than that as they are linked with the catalyst type. (Thecatalysts are typically vanadium pentoxide, titanium dioxide or a noble metal. Catalystshapes include honeycomb plates, parallel ridged plates, rings and pellets.) Forvanadium pentoxide the temperature range is only 600-750°F.5 On the other hand,Cormetech, Inc. (a joint venture by Corning and Mitsubishi, based in Corning, New York)claims that its SCR catalyst has an operating range of 400-850°F and can achieve 90%NOx removal. If the flue gas contains sulfur dioxide, the temperature should be keptabove 608°F.

SCR units are prone to the same adverse reactions that occur with SNCR if theoperating temperatures are above or below the design range. The fouling and pluggingof the catalyst beds associated with poor temperature control led to considerablechanges in the early designs:

♦ Ensuring a mixing time of 0.5-1.0 second between the ammonia injection and thecatalyst bed. This length of time is usually adequate.

♦ Controlling the ammonia injection by monitoring and maintaining the ammonia slip at5 ppm.

♦ Using vanadium and titanium-based catalysts to minimize bisulphate production.

♦ Using titanium and zeolite catalysts to resist sulfur poisoning.

♦ Placing base-metal catalysts downstream of acid-gas control equipment to preventcorrosion by hydrochloric acid. Using catalysts resistant to HCl is another option.

1 McInnes and Van Wormer; Chemical Engineering; September 1990; page 134.2 Exxon Thermal DeNOx Process.3 A. Garg; Chemical Engineering; November 1992; page 128.4 Exxon Thermal DeNOx Process.5 D. Fusselman and D. Lipsher; Oil and Gas Journal; November 2, 1992; page 49.

Page 115: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29101

♦ Installing soot blowing and water wash equipment to remove fouling by fineparticulates.

♦ Using parallel flow catalysts in high dust scenarios in order to minimize plugging anderosion by fly ash. (Honeycomb catalysts are usually installed in low dust unitswhere full advantage of the high surface area can be realized.)

By maintaining ammonia use between 0.9-1.0 mol of ammonia per mol of NOx, SCR canremove 70-90% of the NOx and leave an ammonia slip of only 5-10 ppm.1 This excellentperformance is partially offset by the large capital cost and space required forinstallation. Another potential problem is the narrow temperature range. It may poseproblems trying to place the equipment in the flue gas process flow scheme. Disposal ofthe spent catalyst must also be considered.

10.1.1 NOx Abatement Considerations

The feasibility of NOx reduction technologies is dependent upon many factors:

♦ Does the fuel generate predominantly thermal NOx or fuel NOx? Gas generatesthermal NOx and oil produces mainly fuel NOx. Flue gas recirculation does not affectfuel NOx. Low-NOx burners operate very well with gas firing. With oil firing there isalso good NOx reduction, but the flue gas concentration may still exceed the airquality requirements. This raises the need for post-combustion treatment.

♦ Does the heater have forced-draft or natural draft burners? Boilers are commonlyequipped with forced draft air but heaters often use natural draft. Low-NOx burnersand flue gas recirculation are more common on heaters/boilers with forced-draft airsupply.

♦ What space is available for retrofits? This is especially true for selective catalyticreduction processes, flue gas recirculation and, to a lesser extent, selective non-catalytic reduction.

♦ Can several heaters use a common technology? For instance, selective catalyticreduction could be installed on a common flue gas duct. It is harder with selectivenon-catalytic reduction because of the narrow feasible temperature range, but stillpossible.

♦ Do the existing flue gas temperature profiles lend themselves easily to theinstallation of selective catalytic and non-catalytic reduction? The presence (orabsence) of flue gas waste heat boilers and economizers may make the placementof these selective reduction technologies more difficult.

1 A. Garg; Chemical Engineering; November 1992; page 126.

Page 116: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 102

Figure 10.1: NOx ABATEMENT TECHNOLOGIES1

1 R. Martin and W. Johnson; NOx Control in Fired Heaters; Figures 3, 4 and 5

Page 117: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29103

10.2 SOx

The removal of SOx from flue gas normally involves one of five process categories:

♦ Absorption by non-regenerable solutions

♦ Absorption by regenerable solutions

♦ Adsorption on a solid bed

♦ Direct conversion to sulfur

♦ Direct conversion to sulfuric acid.

The first category is the most predominant and of this category the most commonprocess is wet lime/limestone scrubbing. Limestone has a lower reactivity than lime, butits greater availability and lower cost make it the preferred reagent.

The process is a wet chemical process so there will be some ability to removeparticulates but generally particulates removal is low and plugging can occur. For thisreason, wet scrubbers are usually installed downstream of filters, electrostaticprecipitators or cyclones. This has the added advantage of reducing the volume of wastefor disposal.

Following particulates removal, the flue gas is sent to a prescrubbing step. Wastestreams, preferably containing sodium or calcium reagents, are used at this stage. Themain scrubber ensures direct contact of the flue gas with a lime/limestone slurry.

The scrubber can consist of various technologies. Spray towers are the most efficientfor removing SOx and the pressure drops are low. The main drawbacks are that thescrubbing liquid must be free of solids in order to prevent plugging of the nozzles. Spraytowers are ineffective for removing particles five microns and smaller.

Baffle scrubbers operate at low pressure. Liquid cascades over a series of baffles whilethe gas stream passes through the falling curtain of liquid. Multiple stages are used toachieve the desired desulfurization. Plugging is not a problem and up to 90% ofparticles one micron in size and larger can be removed. Poor liquid/flue gas mixing mayoccur.

In tray scrubbers, the flue gas passes through a perforated tray and then strikes animpingement plate. Even with only one stage, about 90% of the SOx and much of theparticulate matter can be removed. Multiple stages are installed for greater removalrates. The major problems are plugging of the orifices, scale formation on theimpingement plate and the need to have a moveable plate in order to accommodatevarying gas flows. Packed-bed towers have high SOx removal capability but generallycannot achieve good removal of particulates.

After the scrubber, the flue gas passes through a mist eliminator and is vented. Thesulfur-laden liquid waste must be treated using either natural or forced oxidation anddewatered. The resulting calcium sulfate sludge is mixed with lime and fly ash and sent

Page 118: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 104

to disposal. (With the forced oxidation process, the calcium sulfate is often sold asgypsum.) The wastewater must be treated.

Some designs use the dual-alkali process in which the SO2 is removed by a solublealkali such as a solution of sodium sulfite or aluminum sulfate. The solution is thenremoved from the scrubber and reacted with lime or limestone to form insoluble calciumsulfite. The regenerated sodium or aluminum solution is returned to the scrubber.Therefore, loss of the sodium/aluminum reagent is small, the waste is easily disposed ofand there is less scaling and plugging.

Scrubbers using sodium carbonate or sodium hydroxide do not generate solid wastes.The liquids are sent to either wastewater treatment or deep well disposal. The capitalcosts are lower than for calcium-based options, but the operating expenses are higher.Magnesium hydroxide is cheaper than the sodium reagents and is therefore, becomingpopular.

Ammonia scrubbers produce ammonium sulfate, which can be used as fertilizer. HighSO2 removal efficiencies are achieved and there is no, or little, wastewater.

Seawater can be used for SO2 removal because it is slightly alkaline. The flue gas isfirst quenched to about 400°F and then sent to a two-stage seawater scrubber. Theseawater leaving the scrubber is now acidic and is neutralized with fresh sea water. Theeffluent water is then aerated to form sulfates prior to release to the sea. A majordrawback to this process is that the flue gas must be reheated in order to provideadequate plume rise. Obviously, this option is limited to coastal facilities.

Dry scrubbing involves less complex mechanical equipment than wet scrubbing. Theflue gas passes through a spray of lime or soda ash. The heat of the gas evaporates themoisture in the slurry, producing a fine powder of calcium sulfite and calcium sulfate.The particulate matter is collected in a downstream device such as an electrostaticprecipitator or fabric filter. The dry waste is either sent to a landfill or a portion isrecycled back to the spray dryer in order to completely react the reagent.

The most frequent application is for small and medium-sized boilers that use coal with asulfur content less that 1.5%.1 They would thus be satisfactory for many processheaters firing residual oils. The process involves less capital costs then other forms ofwet scrubbing, but operating expenses are higher. Retrofit applications are common.SO2 removal is in the order of 90%, especially if high molal feedrates of reagent areused. Sodium-based reagents (hydroxide and carbonate) are more reactive thancalcium-based reagents (lime and limestone) but are much more expensive.

SO2 removal rates of 50-70% can be achieved by injecting dry sorbent into the furnaceitself or the duct. The most common sorbents are limestone or dolomite(CaCO3/MgCO3). They are injected into the furnace or into the duct upstream of theparticulate removal device.

1 Rounding Up Sulphur; Chemical Engineering; February 1995; page 81.

Page 119: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29105

By humidifying the unreacted CaO from the sorbent injection into the furnace, or byinstalling both furnace and duct injection facilities, a process configuration called hybridsorbent is created. It is capable of achieving 90% SO2 removal.1

As mentioned at the beginning of this section, there are five process categories whenconsidering flue gas desulfurization. The first – non-regenerable absorption - has beendiscussed in some detail since it is the most common category. For more information onother sulfur-removal processes and technologies, the reader is referred to the followingdocuments:

♦ ARPEL Guidelines for the Reduction and Control of Gaseous Emissions fromPetroleum Refineries, Section 10.3.

♦ Rounding Up Sulfur; by V. Kwong and R. Meissner; Chemical Engineering; February1995; pages 74-83. Descriptions of processes specifically related to flue gastreatment are found on pages 80-83.

♦ Sulfur Production Continues to Rise; Chemical Engineering; June 1994; pages 30-35.

10.2.1 SOx and NOx

Early NOx removal catalysts deteriorated in performance when flue gas containing sulfurwas treated. Two reactions occurred. Titanium-based catalysts reacted with SO2 toform SO3, which in turn sometimes led to increased formation of ammonium bisulfate.Alumina-based catalysts reacted with SO3 to form Al2 (SO4)3, which reduced the surfacearea (and activity) of the catalyst.2 Improved, sulfur-resistant catalysts were developed.The results fall into two types:

♦ Processes that remove NOx only, but are resistant to SOx

♦ Processes that simultaneously remove NOx and SOx.

Of the first type, Mitsubishi Heavy Industries, in a joint venture with Corning, calledCormetech, developed a sulfur-resistant NOx Selective Catalytic Reduction catalyst.Details regarding this process/catalyst are provided in the ARPEL Guidelines for theReduction and Control of Gaseous Emissions from Petroleum Refineries, Section 9.1.

There are at least three NOx/SOx removal processes available. These are NOx (NoxsoCorp.),SNOX (Haldor Topsøe ) and Desonox (Lentjes and Lurgi). Very high removalrates are achieved. The Noxso and SNOX processes claim to remove more than 95%of the SOx. Noxso also removes 90% of the NOx. No other performance data wereavailable. More information on these processes can be found in the previouslymentioned Kwong and Meissner article (Chemical Engineering, February 1995) andMajor Stack-Gas Cleanup Process Trial for Ohio in the September 17, 1990 issue of C &EN, pages 35-36.

1 Rounding Up Sulfur; Chemical Engineering; February 1995; page 81.2 High SOx SCR Experience, by M. Yamamura, T. Koyanagi, (both of Mitsubishi Heavy Industries) R.

Iskandar (of Cormetech); no date.

Page 120: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 106

10.3 Particulates

The importance of good particulates removal when desulfurizing flue gas was mentionedin Section 10.2. There are also health reasons for reducing the emission of particulates.Information regarding the types and effectiveness of equipment for removing particulatesis provided in Chapter 12 of the ARPEL Guidelines for the Reduction and Control ofGaseous Emissions from Petroleum Refineries.

Page 121: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29107

11.0 FUEL IMPROVEMENT

The third method of controlling emission from heaters and boilers is to treat the fuel so that it isless likely to produce emissions of NOx and SOx in the first place. Fuel switching is a differentconcept – i.e., the substitution of one fuel for another that is completely different in its emission– producing properties. This has been discussed in Section 9.1. Fuel improvement, in thischapter, means the changing of the fuel’s properties without fundamentally changing the fuelitself.

The question of fuel improvement not only involves heater and boiler operation but it has aprofound influence on the operation and long-term operability of the facility. The decision toimprove fuels used within the facility requires analysis of the future product market (buying orselling), existing and upcoming environmental legislation and the financial resources of thefacility/company.

For a more detailed discussion of this subject, the reader is referred to the ARPEL Guidelineson the Impact of Fuel Switching On Refinery Operations and Atmospheric Emissions.

11.1 Sulfur Removal

The emission factors listed in Chapter 8 clearly highlight the effect of feed sulfur contenton flue gas emissions. Pre-combustion controls involve either blending of high-sulfurfuel with sweeter cutterstock, hydrotreating of the fuel oil and fuel gas sweetening.

11.1.1 Blending

Theoretically sour gas streams could be blended with a sweet gas stream but invariablythe concept is limited to the blending of residual oils with a lighter gas oil, or evenkerosene. No. 6 Oil can be upgraded to No. 5 Oil and even No. 4 Oil. This is doneprimarily to reduce the viscosity and sulfur content. However, emissions of virtually allair contaminants will be improved.

As an example, assume that 1,000 barrels/day of No. 6 Oil at 1.50 wt % sulfur, 0.3 wt %nitrogen and density of 0.96 is upgraded to No. 5 Oil by adding more kerosene (at 0.2 wt% sulfur, 0.02 wt % nitrogen and 0.82 density). The resulting No. 5 Oil has a sulfurcontent of 1.20 wt % and a fuel nitrogen content of 0.236 wt %.

From a yield point of view, an additional 348 bpd of kerosene must be diverted from thediesel (or similar) pool in order to produce the No. 5 oil. This represents a considerableloss in product value. Secondly, the No. 5 Oil is lighter. The heating value will rise fromroughly 17,600 BTU/lb LHV to 17,840 BTU/lb. However, in terms of BTU/barrel, theheating value drops by 2.45%. Therefore, if 1,000 bpd of No. 6 Oil were required for fueloriginally, 1,025 bpd of No. 5 Oil would be required in order to supply the same heatinput. This leaves a surplus of 348 – 25 = 323 bpd of No. 5 Oil which must be sold.Table 11.1 outlines the change in emissions.

Page 122: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 108

Table 11.1: EFFECT OF FUEL OIL BLENDING ON EMMISSIONS, lbs/day

Contaminant No. 6 Oil No. 5 Oil No. 4 Oil

SO2 9,900 8,192 6,101

SO3 126 103 88

NOx 2,181 1,948 1,715

CO 210 215 221

PM 719 432 309

CH4 42 43 2

NMHC 12 12 9

N2O 4.6 4.7 4.8

POM 0.04-0.05 0.06-0.07 0.08-0.09

Formaldehyde 1.0-2.4 1.1-2.4 1.1-2.4

Heavy Metals 6-19 5-15 4-11

PM particulate matterNMHC non-methane hydrocarbonsPOM polycyclic organic matter

If the No. 6 Oil is upgraded to No. 4 Oil, the additional kerosene required as cutterstockis 938 bpd. The sulfur and nitrogen contents are 0.92 and 0.175 wt % respectively. Tosupply the heating needs, 1,051 bpd of No. 4 Oil are consumed leaving a net surplus of887 bpd of No. 4 Oil. The impact upon emissions is shown in Table 11.1.

Blending to a higher grade of fuel oil will improve estimated emissions of all aircontaminants except carbon monoxide, methane in some cases, N20, polycyclic organicmatter and formaldehyde. This may be due to the insensitivity of the published factors torespond adequately to changes in fuel quality. In practice, since No. 5 Oil and No. 4 Oilhave progressively lower viscosity values, there will be better air-fuel mixing andtherefore, more complete combustion. Also, because the sulfur content of the fuel isless, there will be a decrease in the acid gas dew point. This may allow the site toincrease the amount of waste heat recovery from the flue gas. It is however, unlikelythat the overall increase in fuel efficiency will offset the lost product value resulting fromthe downgrading of the additional cutterstock (kerosene in this case).

On the other hand, if air quality standards are stringent and/or disposal of No. 6 Oil isdifficult, fuel blending may be feasible. In most cases, fuel switching to gas is chosen.

11.1.2 Hydrotreating

Hydrotreating involves the use of high pressure hydrogen, at good purity and in thepresence of a catalyst to remove contaminants such as sulfur, nitrogen and metals fromliquid fuels. At the same time, saturation of olefinic and aromatic compounds will occur.

Page 123: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29109

The changing production slates and product quality specifications have placed greateremphasis on this technology. Hydrotreating of heavy gas oils, atmospheric residua andvacuum residua is becoming more common as refiners attempt to improve productquality and upgrade the “bottom-of-the-barrel”.

In terms of fuel for on-site heaters and boilers, hydrotreating of residua is the principalconcern. It is unlikely that middle distillates would be hydrotreated and then used ascutterstock. Hydrotreating the residua reduces the amount of cutterstock that isrequired.

The degree of hydrotreating will depend upon a great many factors such as catalyst typeand activity feed quality, hydrogen partial pressure, reactor temperature and liquid hourlyspace velocity. Processing a variety of crude oils, one refinery achieved the followinghydrotreating results:1

Removal of sulfur 86-91%

Removal of nitrogen 40-52%

Removal of metals (Ni + V) 76-94%

Decrease in residuum density 0.025 – 0.039

Viscosity of product at 122°F 30 – 52.5 SFS

Hydrotreating results in a decrease in the density of the residuum. This is due tosaturation of olefinic and aromatic compounds and cracking of long-chain hydrocarbons.The decrease in density is such that a No. 6 Oil residuum inlet would have the sameoutlet density range as a No. 5 Oil or a No. 5 Oil inlet would have a No. 4 Oil outletdensity. This would greatly reduce the amount of cutterstock. Specifications, as given inASTM D396, show a viscosity range of 45-300 Saybolt Furol seconds (SFS) for No. 6 Oiland 40 SFS maximum for No. 5 Oil. Not all the viscosity values were below 40 SFS2 socutterstock would still be needed when running some crudes or at certain operatingseverity levels. However, the amount of cutterstock would be greatly reduced, and insome cases, eliminated.

The impact on emissions could be estimated using the same procedure as used tocalculate Table 11.1. The effect on the main emissions is shown in Table 11.2. Thistable assumes 1,000 bpd of No. 6 Oil with 1.5 wt % sulfur and 0.3 wt % nitrogen. Thesecond scenario has a hydrotreated fuel oil of No. 5 Oil quality with a sulfur level of 0.18wt % (88% sulfur removal) and a nitrogen content of 0.165 wt % (45% nitrogen removal).Because of the change in density, about 1,025 bpd of the hydrotreated fuel oil isrequired.

1 J. Hohnholt and C. Fausto; Refinery Maintains Optimum Yields via Resid HDS; Oil and Gas Journal;

January 6, 1986; page 66.2 Perry; page 9-6.

Page 124: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 110

Table 11.2: EFFECT OF FUEL OIL HYDROTREATING ON EMISSIONS, lbs/day

No. 6 Oil Hydrotreated Fuel Oil

SO2 9,990 1,229

SO3 126 16

NOx 2,181 1,628

PM 719 432

A comparison of Table 11.1 with 11.2 illustrates the benefits of hydrotreating overblending in terms of heater emissions. Both options will result in improved air-fuelmixing and a lower acid gas dew point, but hydrotreating consumes much lesscutterstock, thereby increasing product revenue. However, hydrotreating does involvesubstantial capital costs and operating expenses in the form of fuel and compressionhorsepower. For more information on this subject, see the ARPEL Guidelines onImpacts of Fuel Switching On Refinery Operations and Atmospheric Emissions, Section7.1.

11.1.3 Fuel Gas Sweetening

Treating of fuel gas will not achieve the dramatic change in emissions brought about bythe blending/hydrotreating of fuel oils, with the exception of reduced emissions of SOx

and CO2. By removing SOx, the acid gas dew point will be reduced, thereby allowinggreater opportunity for waste heat recovery. CO2 removal will improve the heating valueof the fuel gas and reduce the amount of inert material entering the firebox. This willimprove heater efficiency, but only marginally, unless CO2 constitutes a significantportion of the gas, as in the case of pressure Swing Adsorber offgas, which can be asmuch as 40-45 volume % CO2.

11.2 Flare Gas Recovery

Environmental air quality regulations are becoming more stringent. Facilities that burnNo. 6 Oil are faced with the decision of either improving the quality of their residual fuelsor ceasing to use them altogether. A common response has been to use natural gas inplace of fuel oil. This can be costly both in terms of operating expenses but also interms of capital outlay if natural gas pipelines to the facilities must be built.

Flare gas can be considered as a wasted source of internal fuel (and in some cases, aloss of LPG/NGL). It is a necessary component of many oil and gas facilities. Itsprimary function is to relieve excess pressure within the plant or individual processingunits. Although subject to great variation in composition, flare gas is often quite similarto fuel gas, especially if contaminants such as H2S are removed. Under circumstancesoutlined below, flare gas can be potentially recovered and used as fuel. By replacingresidual oil with gas, emissions of greenhouse gases and other air contaminants fromthe heaters and boilers will be reduced. Overall emissions from the plant will also bereduced because flaring will be largely eliminated.

Page 125: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29111

There are three types of flaring. The first type is “incident” flaring. This is characterizedby greatly variable flow rates due to major and minor upsets in operations. It can alsoinclude recurring problems such as passing pressure control valves and safety valves.

The second type of flaring is “background” flaring. In reality, it is composed of themyriad contributors of flaring, such as passing valves. However, in this situation the flowis relatively constant and constitutes the long-term minimum level of flaring. See Figure11.1. At oil batteries, the incoming solution gas, if it is not being recovered for its NGLand natural gas values, can be considered as part of the background flaring.

Figure 11.1: HYPOTHETICAL FLARING PATTERNS

Time Periods

Vo

lum

e o

f F

lari

ng

Major Upset

Minor Upset

Passing Control Valve

Passing Safety Valve

Background Flaring Background Flaring

The third source of flaring is due to maintenance actions. It includes depressuringprocess units or equipment prior to maintenance work and turnarounds. The flare gasvery often contains significant quantities of steam and nitrogen. Whether these inertgases will seriously deteriorate flare gas quality will depend upon the amount ofmaintenance work being undertaken.

When considering flare gas recovery, it is recommended that a flare gas reductionprogram be implemented first. There are two reasons for this:

♦ By reducing the amount of gas being sent to the flare the recovery facilities will besmaller, and therefore, cheaper.

♦ If the flare gas recovery facilities shut down due to mechanical/operating problems,flaring rates will immediately return to their old levels. Experience has shown thatrepair of flare gas recovery facilities can sometimes entail a lengthy wait of weeksand even months for spare parts.

Details of establishing a flare gas reduction program and task force are provided in theARPEL Guidelines for the Reduction and Control Of Gaseous Emissions from PetroleumRefineries, Section 6.2, and the ARPEL Guidelines for the Establishment of a Product

Page 126: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 112

Loss and Waste Accounting and Reduction Programme in Petroleum Refineries,Sections 9.3, 9.4 and 9.5 and Appendix C.

Assuming that a flare gas reduction program has reduced flaring as far as possible, theflaring pattern will look like that shown in Figure 11.1. It will be infeasible to attempt torecover all flaring including that during operational upsets. The equipment would have tobe extremely large and would be operating practically unloaded for the overwhelmingmajority of time. The recovery equipment should be sized for the background level offlaring. Appropriate instrumentation and controls are required to ensure that gas - duringupsets, etc. – is by-passed around the recovery facilities and routed to the flare.

Flare gas recovery has proven to be economically feasible at many sites. The followingpoints must be considered:

♦ Is there a fuel that can be backed out of the fuel pool? Normally, this would beresidual fuel oil. The displaced fuel could also be natural gas, and even fuel gas.Assuming that an existing fuel can be replaced with recovered flare gas, the plantmust be able to dispose of the displaced fuel. For imported fuel it involves thecessation of purchases and for internally-produced fuels and involves the expansionof existing markets.

♦ Is there sufficient long-term background flaring to justify the installation of thefacilities?

♦ Is the flare gas of appropriate quality? For instance, if there is too much nitrogen inthe gas, fuel gas heating values will drop and this may cause firing problems. Ifthere is too much hydrogen there will also be fuel quality problems and the flare gasrecovery compressors may have to be modified.

♦ Will the flare gas require treating? Contaminants such as H2S, cyanides andammonia will have to be removed from the fuel. Other contaminants, such ashydrofluoric acid (from alkylation units) and particulates (from coking units), mayhave be removed prior to recovery facilities or appropriate changes made to themetallurgy of the recovery facilities. New or expanded gas treatment facilities maybe required.

♦ Do sufficient and appropriate burners exist to handle the switch from oil-firing to gas-firing?

♦ Can NGL’s and condensate be feasibly recovered from the flare gas?

Page 127: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29113

12.0 BIBLIOGRAPHY

1. American Petroleum Institute; API Technical Data Book; Washington, DC, USA

2. ARPEL; Guidelines for the Reduction and Control of Gaseous Emissions from PetroleumRefineries; Montevideo, Uruguay; July 1992

3. ARPEL; Guidelines for the Management of Petroleum Refinery Liquid Wastes; Montevideo,Uruguay, July 1992

4. ARPEL; Guidelines for the Management of Petroleum Refinery Solid Wastes; Montevideo,Uruguay, July 1992

5. ARPEL; Impacts of Fuel Switching on Refinery Operations and Atmospheric Emissions;Montevideo, Uruguay; April 1997

6. BP Canada; Variable Operating Pressure for 21-C-6; memo by N. Franklin; Oakville, ON,Canada; December 7, 1976

7. BP Canada; Sidedraw from No. 1 CDU Preflash Column; memo by N. Little; Oakville, ON,Canada; March 8, 1977

8. BP Canada; Preflash Sidedraw; memo by N. Franklin; Oakville, ON, Canada; February 28,1978

9. BP Canada; 11-C-1, Primary Column Simulation; memo by N. Franklin, Oakville, ON,Canada; April 17, 1978

10. British Petroleum; Crude Oil Distillation: Principles and Practice; M. Auckland and E. Horne;Research Project 139; Sunbury-on-Thames; Middlesex, United Kingdom; June 16, 1967

11. Canadian Association of Petroleum Producers; Guide: Global Climate Change VoluntaryChallenge; CAPP Guide to Calculating GHG Emissions; Publication # 1997-0001; Calgary,AB, Canada; April 1997

12. C & EN; Major Stack-Gas Cleanup Process Trial for Ohio; J. Haggin; September 17, 1990

13. Chemical Engineering; Cleaning Up NOx Emissions; R. McInnes and M. Van Wormer;September 1990

14. Chemical Engineering; Make Every BTU Count; A. Garg and H. Ghosh; October 1990

15. Chemical Engineering; Improving Fuel Efficiency with Statistics; T. Mort and I. Verhappen;June 1991

16. Chemical Engineering; Improve Your Fired Heaters; H. Ghosh; March 1992

17. Chemical Engineering; Trimming NOx from Furnaces; A. Garg; November 1992

Page 128: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 114

18. Chemical Engineering; New Horizons in Distillation; J. Humphrey and F. Seibert; December1992

19. Chemical Engineering; Capture Heat from Air-Pollution Control; J. Straitz; October 1993

20. Chemical Engineering; Sulfur Production Continues to Rise; G. Parkinson, G. Ondrey and S.Moore; June 1994

21. Chemical Engineering; Rounding Up Sulfur; V. Kwong and R. Meissner; February 1995

22. Chemical Engineering; NOx control Techniques for the CPI; D. Lambert and T. McGowan;June 1996

23. Chemical Engineering; Advanced Distillation Saves Energy and Capital; F. Lestak and C.Collins; July 1997

24. Chemical Engineering; Catalytic Distillation Extends its Reach; K. Rock, G. Gildert and T.McGuirk; July 1997

25. Chemical Engineering; Distillation Internal Matters; N. Chopey and E. Culp; November 1997

26. Cleaver-Brooks; Boiler Types; http://www.cleaver-brooks.com/Boilersa1.html

27. Cleaver-Brooks; Efficiency Facts; http://www.cleaver-brooks.com/Efficiency1.html

28. Cleaver-Brooks; Glossary; http://www.cleaver-brooks.com/GlossAE.html andhttp://www.cleaver-brooks.com/GlossFP.html

29. Combustion Engineering-Superheater Ltd.; Steam Tables: Properties of Saturated andSuperheated Steam, 3rd Edition; Montreal, PQ, Canada; 1940

30. Exxon Research and Engineering Company; Thermal DeNOx Process; Florham Park, NJ,USA; no date

31. Garcia-Borras, T.; Manual for Improving Boiler and Furnace Performance; Gulf PublishingCompany; Houston, TX, USA; 1983

32. Gas Processors Association; Publication 21-45; Tulsa, OK, USA; 1981

33. Government of Canada, Ministry of Energy, Mines and Resources; Emission Factors forGreenhouse and Other Gases by Fuel Type: An Inventory; Ad Hoc Committee on EmissionFactors; December 1990

34. Government of United States of America, Environmental Protection Agency; Chapter 1.3,External Combustion Sources; Publication 4/93; Washington, DC, USA

35. Hendry, H.; Boiler Efficiency and Testing; John Inglis Equipment Division Publication B 623;Toronto, ON, Canada; no date

Page 129: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29115

36. Hooper, F. and Gillette, R.; Predictive Maintenance of Steam Traps: Combining DemandSide Management and Performance Contracting; http://www.trapo.com/idea-2.htm;presented to International District Energy Association, Indianapolis, IN, USA; June 1995 andupdated October 1997

37. Hydrocarbon Processing; Feedforward Air Control for Fuel BTU Changes; E. Vicknair; July1985

38. Hydrocarbon Processing; Compute Dew Point of Acid Gases; V. Ganapathy; February 1993

39. Hydrocarbon Processing; Understand Boiler Performance Characteristics; V. Ganapathy;August 1994

40. Hydrocarbon Processing; Recover Heat from Waste Incineration; V. Ganapathy; September1995

41. Hydrocarbon Processing; Challenges in Simulating Heat Exchanger Networks; R. Sigal;October 1996

42. Hydrocarbon Processing; Predict NOx from Gas-Fired Furnaces; R. Kunz, D. Smith and E.Adamo; November 1996

43. Hydrocarbon Processing; Controlling Fired Heaters; W. Driedger; April 1997

44. Hydrocarbon Processing; Optimize Fired Heater Operations to Save Money; A. Garg; June1997

45. Hydrocarbon Processing; Revamp Fired Heaters to Increase Capacity; A. Garg; June 1998

46. John Zink Company; Combustion and Industrial Burner Application and Design; Tulsa, OK,USA; no date

47. Makansi, J.; Managing Steam: An Engineering Guide to Commercial, Industrial, and UtilitySystems; Hemisphere Publishing Corp.; New York, NY, USA; 1985

48. Martin, R.; Burner Design Parameters for Flue Gas NOx Control; John Zink Company; Tulsa,OK, USA; no date

49. Martin, R. and Johnson, W.; NOx Control in Fired Heaters; John Zink Company; Tulsa, OK,USA; presented at 1984 Winter National Meeting of American Institute of ChemicalEngineers; March 11-14, 1984

50. Maxim Heat Recovery Equipment, Division of Beaird Industries, Inc.; Heat RecoveryApplication Manual; Shreveport, LA, USA; no date

51. Maxwell; Data Book on Hydrocarbons;

52. Mitsubishi Heavy Industries; High SOx SCR Experience; M. Yamamura, T. Koyanagi and R.Iskander; no date

Page 130: Optimizacion de Combustion ARPEL

The Optimization of Combustion in Boilers and Furnaces

ARPEL Environmental Guideline No. 29 116

53. National Petroleum Refiners Association; Control NOx from Gas-Fired Hydrogen ReformerFurnaces; R. Kunz, D. Smith, N. Patel, G. Thompson and G. Patrick; NPRA Paper AM-92-56; presented at the 1992 NPRA Annual Meeting, March 22-24, 1992

54. Natural Gas Processors Suppliers Association; Engineering Data Book, 10th Edition; Tulsa,OK, USA; 1985

55. North American Mfg. Co.; North American Combustion Handbook, 2nd Edition; Cleveland,Ohio, USA; 1983

56. Oil & Gas Journal; Refinery Maintains Optimum Yields via Resid HDS; J. Hohnholt and C.Fausto; January 6, 1986

57. Oil & Gas Journal Special; Several Technologies Available to Cut Refinery NOx; D.Fusselman and D. Lipsher; November 2, 1992

58. Perry, J., Editor; Chemical Engineers’ Handbook, 4th Edition; McGraw-Hill Book Company;1963

59. Power Specialties; Back to Basics – Steam Traps 101; by D. Fischer;http://www.powerspecialties.com/Armstrong_back_to_basics_Traps101.htm

60. Power Specialties; Energy-Saving Steam Traps Earn Respect at Hüls; by R. Wily;http://www.powerspecialties.com/EnergySavingSteaTraps.htm

61. Sauselein, T.; Stationary Engineering; Business News Publishing Company; Troy, MI, USA;1990

62. Smith, J. and Van Ness, H.; Introduction to Chemical Engineering Thermodynamics, 2nd

Edition; McGraw-Hill Book Company, 1959

63. Sulfur; Leading Burner Designs for Sulfur Plants; January-February 1993

64. Turner, W.; Energy Management Handbook; 2nd Edition and 3rd Edition; Fairmont Press;Lilburn, GA, USA, 1997 (3rd Edition)

Page 131: Optimizacion de Combustion ARPEL
Page 132: Optimizacion de Combustion ARPEL

Mission

It is our mission to generate and carry out activities that will lead to the creation ofa more favorable environment for the development of the oil and natural gas

industry in Latin America and the Caribbean, by promoting:

* The expansion of business opportunities and the improvement ofcompetitive advantages of its members.

* The establishment of a framework to favor competition in the sector.

* The timely and efficient exploitation of hydrocarbon resources and thesupply of its products and services; all this in conformity with theprinciples of sustainable development.

To accomplish this mission, ARPEL works in cooperation with internationalorganizations, governments, regulatory agencies, technical institutions, universities

and non-governmental organizations.

Vision

ARPEL aims at becoming an international level organization that through itsguidelines activities and principles exert an outstanding leadership in the

development of the oil and natural gas industry in Latin Americaand the Caribbean.

Objectives

* To foster cooperation among members.

* To study and assess actions leading to energy integration.

* To participate pro-actively in the process of development of laws andregulations concerning the industry.

* To support actions that expand the areas of activity and increasebusiness opportunities.

* To serve as an oil and gas activity information center.

* To develop international cooperation programs.

* To promote a responsible behavior for the protection of theenvironment, thus contributing to sustainable development.

* To take care of the oil and natural gas industry’s public image.

* To study and disseminate criteria and opinions on the sector’s relevantissues.

Regional Association of Oil and Natural Gas Companies in Latin America and the CaribbeanJavier de Viana 2345 – CP 11200 Montevideo – URUGUAY

Phone: (598 2) 400 6993* Fax (598 2) 400 9207*E-mail: [email protected]

Internet web site: http://www.arpel.org