Vous êtes sur la page 1sur 5

An introduction to velocity-based pore-pressure estimation

COLIN M. SAYERS, Schlumberger Data and Consulting Services, Houston, USA


bnormal pore pressures are encountered worldwide, often resulting in drilling problems such as kicks, blowouts, borehole instability, stuck pipe, and lost circulation. Because of this, a quantitative predrill prediction of pore pressure is required for the safe and economic drilling of wells in overpressured formations. In this paper, the basic concepts used in pore-pressure prediction are defined, and the way in which pore pressure can be estimated using velocity measurements is explained. A predrill estimate of pore pressure can be obtained from seismic velocities using a velocity-to-pore-pressure transform calibrated with offset well data. However, velocities obtained from processing seismic reflection data often lack the spatial resolution needed for accurate pore-pressure prediction, due to assumptions such as layered media and hyperbolic moveout. In the following paper in this section, Chopra and Huffman review the available methods of velocity model building and discuss the advantages and disadvantages of each. Once suitable velocities are obtained, a velocity-topore-pressure transform is required. This article provides the rock physics basis underlying such transforms. Sediment compaction. Following deposition in the marine environment, sediments are initially unconsolidated and have a high porosity and permeability. As a result, the water in the pore space is in pressure communication with the surface, and the weight of the solid phase is supported at the grain contacts and has no influence on the pressure in the fluid (Bourgoyne et al., 1986). The pore pressure p in the fluid is then given by the hydrostatic pressure of a column of formation water extending to the surface (Figure 1). Sediments in which the pore pressure is approximately equal to the hydrostatic pressure are said to be normally pressured, the normal pressure at depth h below the sea surface being given by

Figure 1. In normally pressured sediments, the pore pressure equals the hydrostatic pressure of a column of formation water extending to the surface.

where g is the acceleration due to gravity, and fluid(z) is the fluid density at depth z. As the sediment is buried to greater and greater depth, the weight of the overlying rocks increases, and the increasing stress acting at the grain contacts leads to rearrangement of the grains, resulting in lower porosity and permeability. If the rate of sedimentation exceeds the rate at which fluid can be expelled from the pore space, or if dewatering is inhibited by the formation of seals during burial, the pore fluid becomes overpressured and thus supports part of the overburden load. Overpressure generated in this way is said to result from disequilibrium compaction or undercompaction, this being the most common mechanism for generating overpressure in deepwater sediments. For the purposes of illustration, the total vertical stress S is assumed to be given by the combined weight of the rock matrix and the fluids in the pore space overlying the interval of interest:

Figure 2. Porosity versus depth curves obtained using the relation of Aplin et al. (1995) assuming a grain density of 2.65 g/cc and a fluid density of 1.05 g/cc for * = 0.5, * = 100 kPa and various values of the compaction coefficient .

ported by the fluid pressure p, while the remainder is supported by the rock matrix and is referred to as the effective stress defined by:

Uniaxial strain measurements on unconsolidated sediments have shown that the void ratio e defined by e=/(1) decreases linearly with increase in the logarithm of the vertical effective stress:

where (z) is the density at depth z below the surface, and g is the acceleration due to gravity. Part of this load is sup1496 THE LEADING EDGE DECEMBER 2006

where e* is the void ratio at a chosen effective stress *, and is the compression coefficient (Aplin et al., 1995). This equation may be rewritten as

(1) This equation allows the vertical effective stress to be estimated given a measurement of porosity or density. The pore pressure is then given by the difference between the total vertical stress and the vertical effective stress. Starting from Equation 1, Aplin et al. (1995) have derived a relation describing porosity/depth trends in a horizontally stratified sequence of normally pressured, gravitationally compacted sediments. Figure 2 shows the prediction of this relation for the case in which is assumed to be independent of depth. The vertical effective stress at any depth is then given by Equation 1. In drilling applications, the pressure and overburden gradient, defined as pressure and vertical stress divided by true vertical depth (TVD) below the kelly bushing (KB), are more frequently used than are pressure and vertical stress, and are often expressed in pounds per gallon (ppg) where 1 psi/ft = 19.25 ppg =2.31 g/cc. Thus the overburden gradient (OBG) can be written as (Traugott, 1997)

Figure 3. Porosity/effective stress relation implied by Traugotts relation compared with the prediction of the power law given by Equation 3.

where W is water depth, D is true vertical depth, A is the air gap (vertical distance between KB and sea surface), sea is the seawater density, and avg is the average sediment density between the sea bottom and the depth of interest. Various empirical relations for the overburden gradient have been used. An example is the expression given by Traugott (1997) in which the average sediment density between the sea bottom and the depth of interest is approximated by the expression

where avg has units of ppg, and D, W, and A have units of ft. Since
Figure 4. Density and porosity versus depth below the mudline implied by Traugotts relation compared with the prediction of the power law given by Equation 3.

it follows that this equation corresponds to a sediment density that varies with depth z below sea bottom as where 0 is the density at the mudline. Assuming a grain density s and fluid density f the porosity may be calculated from this equation using the relation (2) Figure 3 plots porosity as a function of effective stress using Traugotts equation and a grain density of 2.65 g/cc and fluid density of 1.05 g/cc together with a best fit of a porosity/effective stress relation of the form (3) where c is a critical porosity above which the effective stress is zero, and max is the effective stress corresponding to =0 in Equation 3. It is seen that the power law given by Equation 3 is in good agreement with the porosity/effective stress relation implied by Traugotts relation. Figure 4 compares the variation of density and porosity versus depth below mudline given by Traugotts relation with the prediction of Equation 3. Good agreement is seen.
DECEMBER 2006 THE LEADING EDGE 1497

Figure 5. Velocity versus porosity obtained from Gardners relation compared with the data presented by Issler (1992) for offshore wells in the Beaufort-Mackenzie Basin, Northern Canada.

Traugotts expression shown in Figure 4. It is seen that the velocity versus effective stress relation that results agrees well with the relation (4) proposed by Bowers (1995). Equation 4 allows the effective stress to be determined from velocity measurements using the relation:

Figure 6. Velocity versus effective stress relation obtained by Traugotts relation with Gardners expression compared with the Bowers relation given by Equation 4.

The pore pressure p can then be determined as the difference between the total vertical stress S and the vertical effective stress . Several other velocity/effective stress relations exist in the literature, and several of these are discussed by Gutierrez et al. in this issue of TLE. The most widely used approach in the industry is the method of Eaton (1975) which estimates the vertical component of the effective stress from the seismic velocity v, via the relation:

Normal and vNormal in this equation are the vertical effective stress and seismic velocity expected if the sediment is normally pressured, while n is an exponent that describes the sensitivity of velocity to effective stress. The pore pressure is then given by:

To use Eatons method, the deviation of the measured velocity from that of normally pressured sediments vNormal must be estimated. Usually, a suitable parameterized expression for vNormal is chosen, and the parameters are obtained by fitting to the shallow velocities assuming that the shallow sediments are normally pressured. However, overpressure often begins at shallow depths thus invalidating this approach. The assumption that shallow sediments are normally pressured is also unnecessary, because the parameters defining vNormal may be determined by an inversion of existing pressure data, as pointed out by Sayers et al. (2002). Consider, for purposes of illustration, the simple linear variation of vNormal with depth given by:
Figure 7. Constant n slice through the 3D space with axes v0, k, and n showing contours of prms defined in the text for n=3.

Pore pressure estimation from velocity. A relation between velocity and effective stress can be obtained from the previous equations relating porosity to effective stress given a relation between velocity and porosity. A suitable relation is Gardners relation (1974) between density and velocity:

where z is depth measured from the seafloor and v0 is the velocity of sediments at the seafloor. Typical values of the vertical velocity gradient k lie in the range 0.6 to 1 s-1 (Xu et al., 1993). To predict pore pressure the parameters v0, k, and Eaton exponent n must be determined. Using a 3D gridbased method (Sayers et al., 2002), the numeric ranges of parameters v0, k, and n may be determined via root-meansquare (rms) analysis of the residuals p = pmeas - ppred:

which allows velocity to be calculated from porosity using Equation 2. Figure 5 compares the velocity versus porosity prediction from Gardners relation, using the parameters given by Castagna et al. (1993) for shales, with the data presented by Issler (1992) for offshore wells in the BeaufortMackenzie Basin, Northern Canada. Figure 6 shows the velocity versus effective stress implied by Gardners equation, with the same parameters as before, and the porosity/effective stress relation implied by
1498 THE LEADING EDGE DECEMBER 2006

where ppred is the predicted pore pressure, pmeas is the measured pore pressure, and N is the number of pore pressure measurements. Figure 7, for example, shows a slice at a constant value of n=3 through the 3D space with axes v0, k, and n. Given an estimate of the error in the pressure measurements used to calibrate the transform, the region of parameter space consistent with the data can be identified, thus allowing a pre-

diction of pore pressure with uncertainty to be made. A similar approach may be used to determine the parameters v0, A, and B in Bowers relation. Discussion. While disequilibrium compaction or undercompaction is the most common mechanism for generating overpressure in deepwater sediments, several other overpressure mechanisms may also occur. Several of these are described below. Clay diagenesis. The smectite-to-illite transformation that occurs as the mixed-layer clay systems in the Gulf of Mexico undergo burial depends strongly on the time-temperature history of the sediment and causes reordering of the clay platelets and redistribution of effective stress (Dutta, 2002). Dutta (1988) incorporates the time-dependent temperature history by use of an effective stress/porosity relation of the form: where 0 is a constant, e is the void ratio defined earlier, A(T) is a polynomial in temperature T, and B(t) is a diagenetic integral depending on time (and temperature) that describes the smectite to illite conversion. Given a relation between velocity and porosity, this equation leads to a velocity/effective stress relation that depends on the timedependent temperature history of the sediment. Unloading. Unloading refers to a decrease in effective stress acting on the rock frame, as may occur if the pore pressure increases in a sediment at a fixed depth. Such an increase in pore pressure may occur, for example, due to temperature increase, clay dewatering or conversion of kerogen to lower molecular weight hydrocarbons. As shown in Figure 8, sediments follow a different path on unloading than on loading (Bowers, 1995), and this difference in the velocity/effective stress relation needs to be taken into account if unloading occurs. As illustrated in Figure 8b, the velocity may drop significantly upon unloading, while the density changes by only a small amount. A velocity/density crossplot may therefore help to distinguish between loading and unloading. Lateral transfer. While the pore pressure within an impermeable shale can increase with increasing depth at a rate that is greater than the hydrostatic gradient, pore pressures within a dipping permeable layer follow the hydrostatic gradient (Figure 9). As a result, flow occurs from the deeper overpressured shales into the sand and from the sand into the shallower shales. As a result of this lateral transfer the sand transmits pore pressure updip (Yardley and Swarbrick, 2000), and this can lead to kicks when drilling into the crest of a dipping sand. A method for calculating the pressure within a dipping sand has

been presented by Stump et al. (1998), and is discussed further by Dutta and Khazanehdari (2006) in this issue of TLE. Conclusion. Several commonly used terms in the pore-pressure-prediction literature have been defined, and various mechanisms that result in overpressure have been discussed. Disequilibrium compaction or undercompaction is the most common source of overpressure in young, rapidly buried sediments, and it is shown how velocity versus effective stress methods arise naturally by coupling the change in porosity with increasing effective stress with the porosity dependence of velocity. In any particular application, additional mechanisms such as clay diagenesis, unloading, and lateral transport may also play a role in causing overpressure to occur, and an appropriate velocity-to-pore pressure transform based on a careful analysis of offset well data together with fit-for-purpose seismic velocities at the proposed drilling location is required for a reliable estimate of

DECEMBER 2006

THE LEADING EDGE

1499

Figure 9. Overpressure resulting from lateral transfer.

Figure 8. Variation of porosity (a) and velocity (b) with effective stress upon loading and unloading.

pore pressure. Uncertainty in the prediction can be reduced by updating the velocity-to-pore pressure transform using data acquired while drilling as described by Malinverno et al. (2004).
Suggested reading. Assessment of , the compression coefficient of mudstones and its relationship with detailed lithology by Aplin et al. (Marine and Petroleum Geology, 1995). Applied Drilling Engineering by Bourgoyne et al. (SPE, 1986). Pore pressure estimation from velocity data: Accounting for pore pressure mechanisms besides undercompaction by Bowers (SPE Drilling and Completion, 1995). Rock physicsThe link between rock properties and AVO response by Castagna et al. (in J. P. Castagna and M. M. Backus, eds., Offset-dependent reflectivity Theory and practice of AVO analysis: Investigations in Geophysics, SEG). Velocity determination for pore pressure

prediction by Chopra and Huffman (TLE, 2006). Fluid flow in low permeable porous media by Dutta (in Migration of Hydrocarbons in Sedimentary Basins, Editions Technip, 1988). Geopressure prediction using seismic data: current status and the road ahead by Dutta (GEOPHYSICS, 2002). Estimation of formation fluid pressure using high-resolution velocity from inversion of seismic data and rock physics model based on compaction and burial diagenesis of shales by Dutta and Khazanehdari (TLE, 2006). The equation for geopressure prediction from well logs by Eaton (SPE paper 5544, 1975). Formation velocity and densitythe diagnostic basis for stratigraphic traps by Gardner et al. (GEOPHYSICS, 1974). Calibration and ranking of pore pressure prediction models by Gutierrez et al. (TLE, 2006). A new approach to shale compaction and stratigraphic restoration, Beaufort-Mackenzie Basin and Mackenzie Corridor, Northern Canada by Issler (AAPG Bulletin, 1992). Integrating Diverse Measurements to Predict Pore Pressure With Uncertainties While Drilling by Malinverno et al. (SPE paper 90001, 2004). Predrill pore pressure prediction using seismic data by Sayers et al. (GEOPHYSICS, 2002). Pressure differences between overpressured sands and bounding shales of the Eugene Island 330 field (Offshore Louisiana, USA) with implications for fluid flow induced by sediment loading: Overpressures in Petroleum Exploration by Stump et al. (Bull. Centre Rech, Elf Explor. Prod., Mem, 1998). Pore/fracture pressure determinations in deep water by Traugott (World Oil, Deepwater Technology Special Supplement, 1997). Some effects of velocity variation on AVO and its interpretation by Xu et al. (GEOPHYSICS, 1993). Lateral transfer: a source of additional overpressure? by Yardley and Swarbrick (Marine and Petroleum Geology, 2000). TLE
Corresponding author: csayers@houston.oilfield.slb.com

1500

THE LEADING EDGE

DECEMBER 2006

Vous aimerez peut-être aussi