Pore Pressure Prediction TLE Nov 2012 Swarbrick - Ikon Science

More documents

Recommendations

Info

Case Studyvelocity—bulk density point according to a suitabledepth interval (e.g., different color each 500m or 1000 feet for example). Increasing burial inthe shallow section is always toward higher velocityand increase in bulk density. Mild development ofoverpressure has the effect of slowing the rate of velocityincrease and bulk density reduction, whereasrapid changes in overpressure may be observed by“reversals” whereby velocity decreases with increasingdepth/vertical stress (but decreasing vertical effectivestress), and bulk density becomes lower atthe same time. Each point remains on or near the“normal compaction curve,” i.e., on trend A (Figure2).Influence of temperature on fluid propertiesTemperature changes fluid properties, in particularreducing density of water and/or hydrocarbonsas well as changing the state of hydrocarbons fromliquid to gas. Because temperatures increase withdepth (at a rate of approximately 30°C/km ± 10°C/km inmost basin settings), pore water has a tendency toward densityreduction and fluid expansion (the compressibility ofwater has the effect of increasing water density but the effectrarely outweighs the temperature effect of density reduction).Barker (1972) first described the fluid expansion of buriedwater and suggested that there would be a significant reductionin vertical effective stress as overpressure was created inthe host sediment. In fact, later research (e.g., Luo and Vasseur,1992; Swarbrick et al., 2002) has shown that the effectis small (< 500 psi even with ultralow-permeability rocks). Interms of volume change, the most significant effect is foundin the generation of methane and other light hydrocarbonmolecules during kerogen maturation (e.g., Meissner, 1976;Ungerer et al., 1983) and during in-situ oil to gas cracking(Barker, 1990).Bowers (1994, 1995) extended the use of velocity-bulkdensity plots to capture the behavior of shale mudrocks during“unloading”—i.e., when effective stress is reduced duringuplift/vertical stress reduction and/or during fluid expansionwhen pore pressures increase from an earlier “maximum”vertical effective stress but without complementary reductionin vertical stress. Because velocity is in part a functionof grain-to-grain contact stress, the effect of vertical effectivestress reduction during fluid expansion is to decrease the shalemudrock velocity. Bulk density, however, will only reduce accordingto the poroelastic behavior of the rock, which is aminor change in porosity. Hence the pathway documentedduring unloading on a velocity-bulk density crossplot is trendB on Figure 2.Figure 2. Velocity-density crossplot illustrating evolutionary pathways withincreasing vertical effective stress (trend A), unloading (trend B), combined porosityreduction/decrease in vertical effective stress (trend C), and chemical compaction(trends D and E).Influence of temperature on rock propertiesBecause traditional pore-pressure prediction relies on predictiverock properties, temperature-driven changes in thegrain-to-grain contact relationship challenge the original assumption,i.e., that porosity attributes can be related to porepressure through the Terzaghi effective stress law. Temperaturecan have a range of effects including:• Precipitation of cements, making the rock stiffer, and/orfilling pore space;• Dissolution of rock grains, ultimately causing collapse ofthe rock structure and reducing pore space;• Weakening of rock grains (more ductile behavior),• Mineral transformation (e.g., smectite to illite; kaoliniteto illite)The full details of chemical diagenesis in shale mudrocksremain poorly known, but reactions such as smectite to illitetransformation (see Lahann, 2001) or kaolinite to illite(Lahann and Swarbrick, 2011a) have been documented inrelation to their influence on vertical effective stress. Both ofthese reactions are kinetically controlled, and occurring overtemperatures in the range 80–150°C (Lahann, personal communication)but where recognition is more typically observedaround 100–110°C. A consequence of these transformationreactions to illite (and potentially other reactions) is that therock becomes more compressible and original grain supporthas been lost—a process referred to as “load transfer” (see Lahannand Swarbrick, 2011b). Many more chemical pathwaysinvolving common minerals in shale mudrocks are possible,which together with the transformations mentioned abovecreate highly unpredictable compaction behavior.Load transfer has the effect of decreasing shale mudrockporosity as minerals (such as illite) are packed more tightly(illite has a preferred orientation of the grains which permitmore efficient packing but which collapse easily when escapeof water allows compaction). The transfer of load, if not accompaniedby expulsion of pore fluid, results in increase inpore pressure and a decrease in vertical effective stress. Thisbehavior is captured on trend C on Figure 2, where a rangeof possible pathways are illustrated to emphasise the multiplecomplementary reactions and load transfers associated withdifferent transformations.November 2012 The Leading Edge 1289
Case StudyFigure 3. Example of data used to determine relationship betweensedimentation rate and fluid retention depth (FRD). In this case,multiple reservoir pressures measurements have revealed a lithostaticgradient-parallel profile of pore pressure, from which the FRD hasbeen determined by intersection with the hydrostatic gradient. Datafrom a wide variety of global basin settings were used to constructFigure 4. Redrawn and modified from Nile Delta data in Mann andMackenzie (1990).Trends in velocity-bulk density have been noted byHoesni (1994) in the Malay Basin in which both velocityand bulk density increase. The interpretation of trend D onFigure 2 is interpreted as chemical compaction of the shalemudrock without reduction of effective stress. It is likely thatthis type of compaction results in pore-filling cements andconcurrent escape of displaced fluids through the connectingpores without development of additional overpressureand simultaneous reduction of vertical effective stress. It isnoted that subsequently, and when bulk densities reach valuesin excess of approximately 2.60–2.65 g/cc, the rate of velocitychange increases, likely reflecting ultralow porosity in theshale mudrock as bulk density approaches rock matrix values(and a minimum porosity). Within this range of densities(i.e., above about 2.60 g/cc) there is no clear discriminationof process (e.g., unloading, load transfer, etc.).Pore-pressure prediction at high temperaturesThe above changes in rock and fluid properties impact verticaleffective stress used with porosity attributes such as sonicvelocity and resistivity to predict pore pressures. Bowers(1995) offers a solution when there is fluid expansion andreduction in effective stress, when solving for velocity, withthe following equation:Velocity = Vo + a[σ max(σ/σ max) (1/u) ] bwhere Vo is taken as 5000 ft/s, a and b are empirical and locallyderived constants, σ is vertical effective stress and u isthe elastoplastic term.The previous equation can be used with confidence wherea clear trend B (Figure 2) can be identified on velocity-bulkdensity crossplots of well data. From the author’s experiencea true “unloading” trend of near-constant density and velocitydecrease is rare. Most trends can best be described as oneof the trend C type profiles shown in Figure 2. There are nopublished solutions for pore-pressure prediction for trend C.Tingay et al., (2009) developed a semiregional correctionto the standard Eaton equation, where the modification ofthe exponent of 3.0 (Eaton, 1975) to a value of 6.5 is requiredto provide a close predictive result when using sonic velocityin a subset of the wells from Brunei. Velocity-bulk densitytrends were not included in this study, only velocity-verticaleffective stress to capture velocity reduction with decrease invertical effective stress along a pseudo-unloading trend. Thissolution is locally calibrated for local use but does not providea relationship which can be applied universally for pore-pressureprediction at elevated temperatures in shale mudrocks.The author has experience of sharp transition zones (rapidreduction in vertical effective stress over a narrow depth interval)in shale mudrocks (and intra-formational reservoirs)accompanied by moderate to severe cementation, identifiedby their ultralow porosity (near matrix bulk density values).Examples include deeply buried, old shale mudrocks such asfound in the Triassic of the Central Graben HPHT play, aswell as young, Miocene sediments in southeast Asia. Bothsequences exhibit no relationship between vertical effectivestress (as inferred from direct pressure tests in interbeddedreservoirs) and porosity. In the case of the Miocene rocks (atdepths in region of 4.0 km and temperatures of 130°C), theporosity is low but vertical effective stress is high, the oppositeto a normal relationship for pore-pressure prediction inyoung, rapidly buried sediments. Yet the shales and reservoirshave high overpressure, created in main by rapid burial. Inthese undrained shale mudrocks, aggressive cementation ismasking any usable prediction relationship with porosity. Incases where there is no usable relationship between porosityattributes (sonic, resistivity, etc.) and effective stress, an alternativegeological-model-based solution is required.Solution using geologic pressure modelDisequilibrium compaction during sediment loading isconsidered the most common cause of overpressure in shalemudrocks (Swarbrick et al., 2002). At elevated temperaturesadditional causes of overpressure such as gas generation andmineral transformations/load transfer may contribute, addingto the existing overpressure created by disequilibriumcompaction during burial. Under these conditions the relationshipbetween porosity and effective stress is either obscureor nonexistent and an alternative to a Terzaghi-relationshipapproach is therefore required. One logical startingpoint for pore-pressure prediction would therefore be an estimateof overpressure in shale mudrocks from disequilibriumcompaction alone, using a method independent of vertical1290 The Leading Edge November 2012
Page 4 and 5: Case StudyFigure 4. Plot of sedimen
Page 6: Case Studydensity crossplots facili

Pore Pressure Prediction TLE Nov 2012 Swarbrick - Ikon Science

Create successful ePaper yourself

Delete template?

Save as template?