Applied Reservoir Simulation Course History Matching
Reservoir Simulation Application Training Course and (Eclipse) Workshop SIS Training and Development, Denver and Houston March 06
Applied Reservoir Simulation Day 2
1
Lecture on Traditional History Matching
March 06
Applied Reservoir Simulation Day 2
2
History Matching and Predictions Simulation of hydrocarbon reservoir performance can be divided into two phases namely:•History Matching: Comprising the adjustment of reservoir parameters in the model until the simulated performance matches the measured information. •Future Performance Prediction: Refers to the simulation of future reservoir performance
March 06
Applied Reservoir Simulation Day 2
3
History Matching • Normally the most time-consuming phase of a simulation study. • Used to demonstrate the validity of the simulation model. • Input historical production rates of oil/gas and well pressures then simulator calculates pressures and secondary products (GOR, WC, etc.) to be compared with history.
March 06
Applied Reservoir Simulation Day 2
4
History Matching Mathematically: Inverse Problem That is, we know the Model and we know the answer, but we do not know the input to the model. There are special techniques for solving inverse problem, but these do not apply to reservoir simulation history matching. Thus, we use trial and ERROR. March 06
Applied Reservoir Simulation Day 2
5
Direct Problem
INPUT Parameters
?OUTPUT?
Inverse Problem
History Matching ?INPUT? ?Parameters?
March 06
System Equations Model Simulator
System Equations Model Simulator
Applied Reservoir Simulation Day 2
OUTPUT
6
History Matching Oil/Gas Field • Special Case of Inverse Problem • Our Geologist, Geophysicist, Petrophysicist, Well Test Engineer give us an estimate of the input parameters (k, kr, Φ, layering, structure, faults, etc.) with an uncertainty level, that is a ± b. • Note sometimes level of uncertainty is not expressly given March 06
Applied Reservoir Simulation Day 2
7
History Matching with Uncertainty Exercise Simplified • Given inverse problem with uncertain input data • Your job: Match the history (within its uncertainty) • Run the prediction • Each group will report their history match parameters and the prediction value March 06
Applied Reservoir Simulation Day 2
8
Variables that can be considered for a history match study •
Porosity (local)
•
Rock Compressibility (Global)
•
Water Saturation (Global)
•
•
Permeability (Local)
Relative Permeability (Global - regional with Justification)
•
Gross Thickness (Local)
•
•
Net Thickness (Local)
Capillary Pressure (Global regional with justification)
•
kv/kh Ratio (Global .. Local?)
•
•
Transmissibility (x/y/z/) (Local)
Mobile Oil Volume (Global..Local?)
•
Datum Pressure (Global)
•
Aquifer Connectivity and Size (Regional)
•
Original Fluid (Global)
•
Pore Volume (Local)
•
•
Fluid Properties (Global)
Well Inflow Parameters (Local)
March 06
Applied Reservoir Simulation Day 2
9
History Matching The following variables are often considered to be indeterminate (high uncertainty): - Pore volume - Permeability - Transmissibility - Kv/Kh ratio - Rel. perm. curves - Aquifer properties - Mobile oil volumes - Well inflow parameters
March 06
Applied Reservoir Simulation Day 2
10
History Matching The following variables are often considered to be determinate (low uncertainty): - Porosity - Gross thickness - Net thickness - Structure (reservoir top/bottom/extent) - Fluid properties - Rock compressibility - Capillary pressure - Datum pressure - Original fluid - Production rates March 06
Applied Reservoir Simulation Day 2
11
Overview of History Matching History matching – Helps to identify the weaknesses in the available data, – Improves the reservoir description and forms basis for the future performance predictions.
March 06
Applied Reservoir Simulation Day 2
12
History Matching The processes which we can control during history-matching are: - The initial quantity and distribution of the fluids within the system - The movement of fluids within the system under pressure gradients
March 06
Applied Reservoir Simulation Day 2
13
Overview of History Matching • Data uncertainties – Oil production rates are usually measured systematically and accurately but Water Oil Ratio and Gas Oil Ratio measurements are done occasionally. – Rate fluctuations are usually smoothed out as they can occur at very short intervals. – Gas rates are not measured accurately especially when part of it is being flared.
March 06
Applied Reservoir Simulation Day 2
14
Overview of History Matching Data uncertainties – Injection data can be less accurate than production data due to measurement errors, fluid losses into other intervals due to leaks in the casing or flow behind the pipe. – Well test data are usually reliable when build up records are taken. – Pressures measured during flow tests are usually less reliable than those obtained during shut in.
March 06
Applied Reservoir Simulation Day 2
15
Overview of History Matching • Reservoir formation parameters are usually measured at the point of observation - wells. – These are porosity, permeability, relative permeability, capillary pressure etc. – Inferred parameters between wells are usually associated with high degree of uncertainty. March 06
Applied Reservoir Simulation Day 2
16
Overview of History Matching • Having set of parameters that match the reservoir performance available does not guarantee accurate future performance predictions. • The duty of reservoir management is to direct efforts in simulation studies to quantify uncertainty in the future performance predictions.
March 06
Applied Reservoir Simulation Day 2
17
History Matching History matching is normally carried out in these stages: - Define which data is determinate and which is indeterminate - Define the data to be compared and the tolerances - Determine the solution technique - Match the pressures (field or reservoir pressure) - Match the GOR/water-cut, etc. - Match the FBHP
March 06
Applied Reservoir Simulation Day 2
18
History Matching Important to know which variables may have a significant effect on the history match and which may not. Thus, at the beginning of the history matching process a sensitivity analysis of the key parameters is suggested.
March 06
Applied Reservoir Simulation Day 2
19
History Matching Before starting to history match: define the data to be matched and the tolerances. Final Match tolerances depend on: - The study objectives - The data quality - The model coarseness - Time and money available March 06
Applied Reservoir Simulation Day 2
20
History Matching There are no “hard and fast” rules for history matching. However, the following may be of some help: - Keep it simple (KISS) - Keep it physical - Use common sense - Don’t panic! - Run model under voidage control until pressures matched, then switch to principal product rate - Proper use of pseudos is more consistent than rock curves - Correct the problem, not the symptom - The most difficult problems may require a rethink of the base data March 06
Applied Reservoir Simulation Day 2
21
History Matching Pressure match - Run the model under reservoir voidage control - Examine the overall pressure levels, adjust the pore volume/aquifer properties to match overall pressure - Match the well pressures, modify local PVs/aquifers to match overall pressures; modify local transmissibility to match pressure gradient March 06
Applied Reservoir Simulation Day 2
22
History Matching Saturation match - Normally attempted once pressures matched - Most important parameters are rel. perm. curves and permeabilities - Try to explain the reasons for the deviations and act accordingly - Changes to rel. perm. tables should affect the model globally - Changes to permeabilities should have some physical justification - Consider the use of well pseudos - Assumed layer KH allocations may be incorrect (check PLTs, etc.) March 06
Applied Reservoir Simulation Day 2
23
History Matching Well PI match - Not usually matched until pressures and saturations are matched, unless BHP affects production rates - Must be matched before using model in prediction mode - Match FBHP data by modifying KH, skin or PI directly
March 06
Applied Reservoir Simulation Day 2
24
History Matching Automatic history matching does not exist. New Eclipse SimOpt program is a history matching aid. Try to minimize an objective function over the time period of the history: n
Q=∑R where
with
i =1
2 i
Ri = Wi ( X − X o )i X = calculated data Xo = actual data Wi = weighting factor
March 06
Applied Reservoir Simulation Day 2
25
History Matching Adjustments Simulation field pressure too high
Possible Changes Pore Volume? Aquifer? Oil Initially in Place (s, So) Energy? Gas cap size
March 06
Applied Reservoir Simulation Day 2
26
History Matching Adjustments Break Through: OK, After BT WC too high
Possible Changes Krw / Kro ratio decrease? Aquifer size?
March 06
Applied Reservoir Simulation Day 2
27
History Matching Adjustments Water BT too early, After BT slope - OK
Possible Changes Effective end point Krw ? Horiz. Permeability of well to aquifer layer? Shale or barrier between wells and water? Vertical permeability between wells and water? Numerical dispersion / grid effect?
March 06
Applied Reservoir Simulation Day 2
28
History Matching Adjustments Gas BT simulation very late, After BT slope OK
Possible Changes Effective Krg end point? Shale or barrier effecting coning? Vertical permeability between wells and gas? Numerical dispersion / grid effect?
March 06
Applied Reservoir Simulation Day 2
29
History Matching Adjustments Gas BT OK, After BT simulation slope in error
Possible Changes Krg / Kro ratio increase? Supply of gas?
March 06
Applied Reservoir Simulation Day 2
30
Predictions Most simulation models are used to produce production forecasts. Exact details of prediction cases depend on particular field, so difficult to generalize. - During predictions, well rates usually controlled primarily by FTHP/FBHP - Most simulators can model a wide variety of operating constraints - Check that the transition from “history” to “prediction” is smooth - Monitor carefully “automatic” workovers - Take care when using the model to plan infill drilling - Simulation model recoveries are usually optimistic March 06
Applied Reservoir Simulation Day 2
31
History Matching Restart Runs
Field production rate
History Period
Prediction Period
Base Run
Restart Run Present day
March 06
Applied Reservoir Simulation Day 2
Time 32
End of Lecture 2
March 06
Applied Reservoir Simulation Day 2
33