AN INTRODUCTION TO PRESSURE TRANSIENT TESTING OF INJECTION ... 2017_02_22... · an introduction to...
Transcript of AN INTRODUCTION TO PRESSURE TRANSIENT TESTING OF INJECTION ... 2017_02_22... · an introduction to...
AN INTRODUCTION TO PRESSURE TRANSIENT TESTING OF INJECTION WELLS AND USE FOR UIC PERMIT AND LANDBAN PETITION COMPLIANCE
GROUNDWATER PROTECTION COUNCIL ANNUALUNDERGROUND INJECTION CONTROL CONFERENCEFebruary 2017; Austin, Texas
Ken CooperLewis Wandke & Aaron Payne
PETROTEK ENGINEERING CORPORATION
Pet
rote
k - 2
017
OBJECTIVES
Overview of Pressure Transient and Fall Off Testing
Why Conduct Testing?
How Does Fall Off Testing Work?
Fall Off Analysis and What Can it Tell Us?
Critical Issues for Analysis and Compliance Use
Suggestions for Optimizing Fall Off Tests
Summary and Questions
2
2017 GWPC UIC Conference
Pet
rote
k - 2
017
WHAT IS PRESSURE TRANSIENT TESTING?
A well test that involves recording pressure versus time to determine how flow rates influence pressure behavior measured in a well
Mathematical relationships between flow rate, pressure and time are applied to data to infer properties and conditions of the well and reservoir
3
2017 GWPC UIC Conference
Pet
rote
k - 2
017
MAIN TYPES OF PRESSURE TRANSIENT TESTS Fall-off or Build-upstable flow period followed by pressure recovery period after the shut-in of a tested well
Drawdown or Injection (Single or Multi-rate) pressure decrease/increase during stable test well flow periods
Step Rateinjection pressure increase versus time for multiple, consecutive, constant rate, equal duration steps
Interference (Standard or Pulse)observation (test) well pressure response due to rate changes in offset active well
4
2017 GWPC UIC Conference
Pet
rote
k - 2
017
WHY PERFORM WELL TESTING? Reservoir characterization over a larger scale around a well
than logs or cores can investigate
Real-world field confirmation of well capacity and pump pressure requirements
Assessment of current well condition (completion efficiency from borehole into disposal reservoir, aka skin factor)
Evaluate fracture pressure
Determine reservoir continuity - pressure interference between wells and inter-well/directional properties
5
2017 GWPC UIC Conference
Pet
rote
k - 2
017
WHY PERFORM WELL TESTING?(REGULATORY)
Ambient monitoring requirements (40 CFR 146.13 & 146.68)
Investigate permeability-thickness and reservoir extent assumptions used for cone-of-influence calculations
Verify that well conditions remain consistent over time with values used as the basis for regulatory approvals
Provide insight regarding reservoir pressure trends for comparison to model projections in permits and/or no-migration demonstrations
6
2017 GWPC UIC Conference
Pet
rote
k - 2
017
WHY PERFORM WELL TESTING?(TECHNICAL)
Ongoing understanding of realistic well capacity, operating limitations, changing conditions, and well life expectancy
Aid with differentiation of reservoir limitation or wellbore conditions as potential reasons for decreasing capacity
Provide insight into wellbore plugging for the evaluation of treatment options (timing and near wellbore or deep damage)
Provide insight regarding reservoir pressure trends for comparison to projections that might impact migration (manage liability)
7
2017 GWPC UIC Conference
Pet
rote
k - 2
017
HOW DOES FALL OFF TESTING WORK? Pressure propagates in a reservoir as a log function of time and
distance from the source term (well)
8
2017 GWPC UIC Conference
Pet
rote
k - 2
017
HOW DOES FALL OFF TESTING WORK? A rate change at a well will cause a pressure change in the reservoir
that can be measured in a well
9
2017 GWPC UIC Conference
Pet
rote
k - 2
017
HOW DOES FALL OFF TESTING WORK? The pressure change is represented by the superposition of new shut-in
rate, q = 0, on the prior injection rate
10
2017 GWPC UIC Conference Courtesy KAPPA (modified), 2007
Pet
rote
k - 2
017
HOW DOES FALL OFF TESTING WORK? Must account for changes from surface to bottom hole and for the
transition from borehole to porous media
11
2017 GWPC UIC Conference
Pet
rote
k - 2
017
HOW DOES FALL OFF TESTING WORK?
Exponential Integral (Ei) solution to Diffusivity Equation
Log approximation
Assumes radial flow with homogenous & isotropic conditions
More complicated scenarios require more complex treatments and introduce analysis uncertainty
Differential equations used to represent pressure behavior in a porous media also used to evaluate heat transfer and electricity
12
2017 GWPC UIC Conference
Pet
rote
k - 2
017
BASIC FALL OFF ANALYSIS EQUATIONS
Δp = -162.6 q µ B [ log k t – 3.2275 + 0.869 s ]k h Φ μcrw
2
m = -162.6 q µ Bk h
s = 1.1513 [ P1hr – Pwf – log ( k ) + 3.2275 ]m Φ μcrw
2
Earlougher (1977), Matthews and Russell (1967)
13
2017 GWPC UIC Conference
Pet
rote
k - 2
017
WHAT IS SKIN FACTOR?
A mathematical convenience related to equivalent wellbore radius used to represent wellbore performance as compared with an ideal completion
14
2017 GWPC UIC Conference Courtesy Schlumberger (modified), 2006
Pet
rote
k - 2
017
FALL OFF TEST ANALYSIS
Plotting and processing these changes as various functions or history-matches of observed data to idealized predictions allows us to infer well and reservoir properties
Diagnostic dp/derivative log-log plot
Semi-log plots
Complex functions and superposition
15
2017 GWPC UIC Conference
Pet
rote
k - 2
017
WHAT DOES AN IDEAL FALL OFF TEST LOOK LIKE? 16
2017 GWPC UIC Conference
Pet
rote
k - 2
017
WHAT DOES AN IDEAL FALL OFF TEST LOOK LIKE? 17
2017 GWPC UIC Conference
Pet
rote
k - 2
017
FLOW REGIMES – PATTERN RECOGNITION Near wellbore, reservoir or boundary conditions?
18
2017 GWPC UIC Conference Courtesy Fekete (modified), 2009
Early Time Middle Time Late Time
Log
Pres
sure
Cha
nge
and
Der
ivat
ive
(rate
of C
hang
e)
Log Time
Pet
rote
k - 2
017
ASSUMPTIONS MATTER:FALLOFF TESTING UNCERTAINTY Total Historical Injection Volume (Pseudo Time, Tp)
Rates and Pre-test Flow Period Duration
Year to Year Operating Changes
Offset Injection
Mobility (k/u)inner/(k/u)outer or Transmissivity (kh/u) with Distance
Reaching Heterogeneity or Boundary
19
2017 GWPC UIC Conference
Pet
rote
k - 2
017
WHAT ASSUMPTIONS MATTER?FALLOFF TESTING: HISTORICAL VOLUME
20
2017 GWPC UIC Conference
could be confused with changing pressure, skin, or permeability
Base CaseFix Pi andattempt analysis
1/10 of historicalvolume, match Pi
Pet
rote
k - 2
017
WHAT ASSUMPTIONS MATTER ?FALLOFF TESTING: MOBILITY, RADIAL FLOW?
21
2017 GWPC UIC Conferencecould significantly change interpreted permeability, skin and pressure
Pet
rote
k - 2
017
WHAT ASSUMPTIONS MATTER ?FALLOFF TESTING: OFFSET INJECTION
22
2017 GWPC UIC Conferenceeasily confused with altered permeability (5000 mdft) or pressure (Pi = 2000)
Pet
rote
k - 2
017
PRIMARY REGULATORY USE AND LIMITATIONS OF FALL OFF TEST RESULTS
Changing kh or skin factor with time? Operating pressure increase Reduced injection capacity Is it different operating conditions? Did the effective thickness or viscosity change?
Accelerated reservoir pressure increase? Not P*, not Pwf, but final shut-in pressure or Pave
23
2017 GWPC UIC Conference
Pet
rote
k - 2
017
FALL OFF TEST ANALYSIS – COMPLIANCE USE
Differences between Pi, Pwf, P*, Pave, P1hr
Pwf is “flowing” pressure impacted by near wellbore effects, skin, well geometry, friction, density, viscosity
P* aka “false pressure” only = Pi in an ideal infinite acting reservoir, limited use and requires corrections for reservoir geometry
P1hr is extrapolation from radial-flow slope “m”, used for calculations and is not 1-hour gauge value
P* ≠ Pwf ≠ P1hr ≠ Pave ≠ Pi
24
2017 GWPC UIC Conference
Pet
rote
k - 2
017
FALL OFF TEST RESULTS – COMPLIANCE USE
History Match“Static” BHPRise at End ofFall Off Tests < Simulation
25
2017 GWPC UIC Conference
Pet
rote
k - 2
017
FALL OFF TEST RESULTS – COMPLIANCE USE
History Match…..What aboutYears of BHP from Fall OffPwf > Model ?
26
2017 GWPC UIC Conference
Pet
rote
k - 2
017
FALL OFF TEST RESULTS – COMPLIANCE USE
History Match….. After WellStimulation,BHP of Fall OffPwf < Model
27
2017 GWPC UIC Conference
Pet
rote
k - 2
017
FALL OFF TEST RESULTS – COMPLIANCE USE
Skin and kh/uWith Static dPExplain Pwf ..…Near WellborePlugging thatwas Reversible
28
2017 GWPC UIC Conference
Pet
rote
k - 2
017
FALL OFF TEST RESULTS – COMPLIANCE USE Together with other operating and well data, fall off trends should
be examined together, in context
An isolated viscosity increase one year might yield higher Pwf and apparent skin, without changing deep reservoir kh/u
Early in a well life, fall off testing could “see” a boundary in a late time fall off test slope before significant deviation from simulation pressure predictions
Offset well operation in a reservoir could cause Pwf and Pavemeasured in a fall off test to rise without any associated skin factor increase or decreasing transmissivity
29
2017 GWPC UIC Conference
Pet
rote
k - 2
017
DESIGN ISSUES - OPTIMIZE FALL OFF TESTING Larger skin, smaller permeability, changing storage will extend the test
duration needed to see useful radial flow period
Increasing rate = increased semi-log slope (m), important in high permeability systems or with lower gauge resolution
Constant injection rates and periods help but are not critical using newer simulators and superposition techniques
Pay attention to offset well rates, it can be easy to “see” reservoir issues or artificial slopes that are actually interference effects
It is hard to differentiate between rock and fluid heterogeneity (kh/u) so understand fluid viscosity, density and temperature
30
2017 GWPC UIC Conference
Pet
rote
k - 2
017
FIELD PRACTICE - OPTIMIZE FALL OFF TESTING Use BHP gauges for highly variable density, significant tubing friction or
when positive wellhead pressure will not be maintained
If valves do not seal, injection continues after “shut-in”
Gauge position close to completion top is often better
Run gradient surveys in/before and out/after testing
Pressure without depth is of limited use, verify depths in & out
Specify KB or GL
psig or psia (get gauge calibrations)
31
2017 GWPC UIC Conference
Pet
rote
k - 2
017
ANALYSIS ISSUES - OPTIMIZE FALL OFF TESTING Assuming “good” pressure data is obtained, useful pressure transient
analysis is not possible without good rate data; garbage q in = garbage kh/u out
If Pi, Pwf and P* are unrealistic in a graphical or simulation match; the interpretation could be misleading
Wellbore transients typically dominate reservoir transients
Simple models that acknowledge uncertainty can sometimes provide more insight than overly complicated approaches that are not justified nor unique
Fall off tests must be used in context with all available information
32
2017 GWPC UIC Conference
Pet
rote
k - 2
017
Contact Questions and Discussion33
Ken Cooper, MS, PEPrincipal, Engineering ManagerLewis Wandke, PE; Aaron Payne, MS, PG
Petrotek Engineering Corporation5935 South Zang Street, Suite 200Littleton, Colorado 80127(303) 290-9414 x 415
www.petrotek.comall rights reserved
2017 GWPC UIC Conference
Pet
rote
k - 2
017