Study of Hot-air Recirculation around Off-road Tier-4 Diesel Engine Unit

Using CFD Siddharth Jain, Yogesh Deshpande, Atul Bokane, Arun Kumar Santharam, and Deepak Babar,

Halliburton

Abbreviations: Computational fluid dynamics

(CFD), boundary condition (BC).

Keywords: Hot-air recirculation, Tier-4 diesel

engine, recirculation zone, aerodynamic deflectors

Abstract

The goal of this study was to increase

the cooling efficiency of Tier 4 diesel engine

installations on well service equipment. These

efficiency improvements were directed at

reducing hot-air recirculation around the

engine’s radiators. Well service equipment units

are designed for use in upstream oilfield

activities. These units are powered by diesel

engines that are required to comply with the new

emission standards defined by the United States

Environmental Protection Agency (EPA) for off-

road engines. The new Tier 4 emission

standards have led engine manufacturers to

increase the cooling requirements for engine

packages, thus increasing the burden on

existing cooling systems. A traditional radiator

uses ambient air and is influenced by

parameters such as air temperature, wind

direction, wind speed, and nearby objects. This

equipment is intended to work in operating

temperatures greater than 115°F but are

currently limited in ambient capability.

Particularly during the summer, these engines

are prone to overheating. One cause of radiator

inefficiency is the recirculation of hot air around

the radiator, which significantly reduces cooling

system performance.

Computational fluid dynamics (CFD)

modeling provides an effective means to assess

the cause of recirculation and offers a way to

evaluate solutions to improve unit performance.

CFD modeling was performed, and based on the

results, recirculation zones were identified. For a

particular recirculation zone, several

combinations of aerodynamic deflectors were

designed to deflect hot air and prevent

recirculation. Analysis was conducted using

ANSYS FLUENT V13 software. The geometry

creation of the computational domain was

performed using SolidWorks 2011 software. The

computational domain includes the radiator, fan,

front cab, fuel tank, and simplified engine to

study the effect of hot-air recirculation.

Considering the effect of the radiator

and fan, CFD analysis provides results in the

form of velocity vectors and path lines, which

provide actual flow characteristics of air

circulation around the radiator. The CFD results

were in excellent agreement with the data

measured during physical testing. The CFD

results indicated that adding deflectors would

greatly reduce recirculation and improve unit

performance, enabling the units to operate in

higher ambient conditions. The output of this

analysis is intended to be used at field locations

across the globe for reliability in operation.

Introduction and Background

Recirculation has been an important

concern during the last few decades in the oil

and gas industry. General rules for predicting the

occurrence of hot-air recirculation have been

around for a considerable time [1, 2]. Hot-air

recirculation occurs where the use of a heat

transfer device is required, such as at power

plants or in the oil and gas industry [1]. The most

universal equipment used for cooling purposes

in the oilfield service industry is radiators.

Radiators are effective heat transfer devices,

but, because they use ambient air, their

performance is influenced by air temperature,

wind direction, wind speed, and by the proximity

of other air coolers and environmental flows. In

particular, radiators can experience recirculation

of hot exhaust air back to the air intake side,

significantly reducing the overall unit

performance.

With the expressive computer capability

and extensive development in the simulation

field, CFD has drawn attention in recent years.

CFD models, if created correctly, can account

for the complex interactions between the

ambient conditions (wind speed and air

temperature) and equipment and heated plumes

that are often the cause of recirculation

problems. CFD models have been used to

evaluate recirculation zones and also to design

deflectors to improve overall cooling

performance. The models provided sufficiently

accurate predictions over a range of operating

conditions, which were not possible using other

methods. With recent advances in computational

speed and modeling capabilities, the complex

three-dimensional geometries of the equipment

can now be modeled with only minor

simplifications.

Problem Definition

Every off-road diesel unit must comply

with new emission standards as defined by the

United States EPA. According to the EPA, new

equipment built after Jan 1, 2011 must

incorporate diesel engines that comply with the

new Tier 4 standard. Specific diesel Tier-4 units

are designed for use in upstream oilfield

activities (Fig. 1). The unit discussed was

intended to operate in the southern region of the

U.S. where temperatures can reach 115°F

regularly. Because this unit experienced air

recirculation problems it was only capable of

operating at temperatures below 104°F without

engine overheating. This was the result of

radiator inefficiency caused by recirculation of

hot air around the Tier-4 unit and, thus, a

significant reduction of cooling system

performance.

Fig. 1: Tier-4 Unit.

Methodology

CFD modeling provides an efficient way to

evaluate the cause of recirculation and offers an

approach to estimate keys to enhance unit

performance. For this study, CFD analysis has

been divided into three stages.

1. CFD of the fan. 2. CFD of the total computational domain. 3. Design optimization for the deflector.

Figs. 2 and 3 show the computational

domain for the fan and complete Tier-4 unit,

respectively. Initially, the fan performance was

evaluated by validating its speed and direction of

rotation. The complete domain was considered

to gain comprehensive information about

recirculation patterns around the engine unit.

Fig. 2: Computational Domain—Fan.

Fig. 3: Computational Domain—Complete Unit.

CFD Setup

Fig. 4 shows a typical simulation

parameter used in ANSYS-FLUENT13.0. For

modeling the fan, inlet pressure and outlet

pressure were defined as boundary conditions

(BCs). The speed (measured in rpm) of the fan

was entered as an input.

For the complete computational domain

[fan + engine + radiator], BCs were set up for

inlet velocity, inlet and outlet pressure, and the

wall conditions. The input was provided in terms

of velocity of air flow, pressure, fan rpm, and

radiator resistance.

Fig. 4: Simulation Parameter.

Fig 5 shows meshing performed using

an ANSYS workbench patch independent

meshing option. It has 3.5 million tetrahedral

elements to discretize computational domain.

The quality of mesh was assured to be

acceptable.

Fig. 5: Meshing—Computational Domain.

Turbulence Model – K-Є turbulence model

with standard wall function

Temperature – 298 K

Pressure – Atmospheric pressure

Flow media – Air

3D, Pressure Based Implicit, steady

Solution Method – SIMPLE, First Order

Upwind

Convergence Criteria – 1e-4

Meshing – Extracting volume method with

tetrahedral elements.

Deflector Locations and Design Profiles

The deflector’s location was decided

after identifying the recirculation zone around the

radiator. The deflectors were positioned

• In between the front cab and

radiator

• On the bottom of the radiator

• On the side of the radiator (side

deflector)

Fig. 6 shows the locations for various

deflectors, which were placed on the Tier-4 unit.

Fig. 6: Deflectors Location on Tier-4 Unit.

Fig. 7 shows the design profiles of the

various deflectors. CFD analysis was performed

for the following designs:

• Straight deflectors.

• Inclined deflectors.

• Curved deflectors

Fig. 7: Deflector’s Designs Profile.

Results

CFD analysis was performed for all

three design profiles. The optimized profile of the

deflectors was selected based on the hot-air

recirculation zone and path line pattern. The

straight and inclined deflector showed the hot air

was again reversed back to the radiator inlet,

and the recirculation zones formed at the bottom

of the fuel tank. These results were not desirable

to control hot-air recirculation. However, with the

curved deflector, due to its aerodynamic nature,

most of the air contacts the curved trajectory,

which directs the hot air flow away from the

radiator and engine inlet. This improved the unit

performance significantly. It was decided to use

the curved deflector to limit hot-air circulation

problems around the Tier-4 engine. The length

and size of the curved deflector was designed as

per material availability and manufacturability.

Figs. 8a, 8b, and 8c clearly show the

recirculation of air around the Tier-4 unit without

any deflector. The velocity vector showed the

flow of hot air was forming recirculation zones at

the radiator inlet and around the engine, which

caused engine overheating. Eventually, this

resulted in poor unit cooling performance.

Fig. 8: (a) Velocity Vectors—Domain without

Deflectors.

Fig. 8: (b) Velocity Vectors—Domain without

Deflectors.

Fig. 8: (c) Velocity Vectors—Domain without

Deflectors.

On the basis of CFD analysis with the

base domain without a deflector, several

recirculation zones were identified. The curved

deflector was designed in succession over

straight and inclined deflectors.

It was observed that, once deflectors

were placed at their respective positions, the

recirculation of the hot air was limited, thereby

improving the reliability of the unit. Figs. 9a

through 9d show the flow around the radiator,

and it is apparent from the velocity vectors that

hot air was deflected away from the radiator and

engine unit by the deflector.

Fig. 9: (a) Velocity Vectors—Domain with

Deflectors.

Fig. 9: (b) Velocity Vectors—Domain with

Deflectors.

Fig. 9: (c) Velocity Vectors—Domain with

Deflectors.

Fig. 9 (d) Velocity Vectors—Domain with

Deflectors.

Conclusions

The hot-air recirculation around the Tier-

4 unit was predominantly causing poor cooling

performance and frequent potential engine

overheating conditions. CFD was an effective

and comprehensive tool to evaluate this

phenomenon by considering a larger

computational model of the entire Tier-4 unit and

its surrounding components.

The base computational model with the

Tier-4 unit evaluated flow characteristics in

terms of flow direction of hot air, path lines, and

velocity vectors. The locations and different

configurations of deflectors to limit the hot-air

recirculation toward the radiator and engine inlet

were identified.

Various designs of deflectors, including

straight, vertical, inclined, and curved, were

studied to reduce the hot-air recirculation in the

Tier-4 units. It was determined that the curved

deflector was more effective and provided the

best solution by restricting the hot-air

recirculation.

The key and important outcomes of this

study are identified below.

• Recirculation of hot air occurring around

a Tier-4 unit was evaluated using

ANSYS FLUENT 13.

• ANSYS FLUENT 13 has excellent

meshing capability for complex and

large domain volume.

• CFD results were presented in the form

of velocity vectors and path lines, which

provide actual flow characteristics of air

circulation around the radiator.

• Several combinations of deflectors were

designed as straight, inclined, and

curved deflectors to limit hot-air

recirculation around a Tier-4 unit.

• Particularly, the curved deflectors

showed significant improvement to

control hot-air recirculation in

comparison to straight, vertical, and

inclined deflectors.

• The CFD results were in excellent

agreement with the data measured

during physical testing.

• Adding deflectors would reduce

recirculation and improve performance,

enabling the units to operate in severe

conditions.

Acknowledgement

The authors thank the management of

Halliburton for their support and permission to

publish this paper. They also express gratitude

to all of the team members who contributed to

the job design, preparation, and execution of the

operation to achieve the results presented in this

paper.

References

1. J.A. Rojers, K. Won, and W. Stang.

Validation of CFD Models for Evaluating

Hot-Air Recirculation in Air- Cooled Heat

Exchangers. Paper presented at the

AlChE Spring Meeting. Houston, Texas,

14–18, March 1999.

2. A.Y. Gunter and K.V. Shipes. Hot Air

Recirculation by Air Coolers. Paper

presented at the Twelfth National Heat

Transfer Conference AlChE—ASME.

Tulsa, Oklahoma, 15–18, August 1971.