Pressure Index (PI): A Simple Novel Method to Normalize Formation Pressure in Offshore Malaysia

Pressure Index (PI): A Simple Novel Method to Normalize Formation Pressure in Offshore Malaysia

Yudiyoko Ega Sugiharto*

*PETRONAS Carigali SDN BHD, Kuala Lumpur, Malaysia

ABSTRACT

Pressure analysis is concerned with the study of systematic variations of reservoir pore pressure with depth. The most common interpretation for pressure analysis is pressure-depth plot analysis, but other techniques that magnify understated pressure differences are also available. The measurement of formation pressure is of immense value in quantitative evaluation and prospect risk. Once the pressure data has been acquired, we need to understand how to interpret the data received because reservoir pressure data has numerous applications and misinterpreting it could make the results misleading. At equilibrium state (i.e. there are no net forces and no acceleration), a fluid in the system is called hydrostatic equilibrium. Hydrostatic pressure increases with depth measured from the surface due to the increasing weight of fluid exerting downward force from above. The traditional pressure evaluation is usually done in conventional unit such as psi, kPa, psi/feet, psi/m, kPa/m, ppg. The current work will introduce the concepts and definitions of formation pressure evaluation using Pressure Index (PI) with the unit g/cc. For better understanding of the application of PI, some reservoir studies are also discussed in this paper.

Introduction

In oil and gas industry it is very common to use psi as reservoir pressure unit. Pressure condition (i.e hydrostatic, under pressure or over pressure), can be obtained by plotting pressure against true vertical depth which called pressure profile analysis. Howes (1999) stated the typical applications of pressure profile analysis guide us to (a) identification of pore fluid type, (b) estimation of pore fluid properties, (c) estimation of fluid contacts and hydrocarbon column heights, (d) quantification of depletion and overpressure, (e) identification of hydrodynamic conditions, and (f) investigation of lateral and vertical reservoir continuity and connectivity.

Figure 1 describes the differences between lithostatic, abnormal, subnormal and normal pressure based on range of PI.

Howes (1998) mentions that, as the rule of thumb, normal hydrostatic gradients range between 0.433 psi/ft and 0.465 psi/ft. As best practices definition in petroleum industry, overpressure is referring to pressure higher than normal hydrostatic pressure which required higher mud weight to keep a well from flowing while drilling and subnormal pressure refers to pressure lower than hydrostatic pressure or depleted due to post-production.

In this paper we manage to convert pressure to the density units (gram/cc) and use these as an alternative methods of pressure profile analysis and calling it Pressure Index (PI).

The Relationship of Hydrostatic Pressure and Fluid Density

The density of any material is calculated by dividing the mass of the matter by the volume of the matter.

Hydrostatic pressure depends on the density of the liquid and liquid column as per Equation (1) below:

………………………………………………………………….       (1)

By re-arranging Equation (1), Pressure Index (PI) can be estimate by knowing pressure and liquid column:

……..…………………………………………………………….   (2)

is not only known as mud weight equivalent (EMW) or Specific Gravity (SG) by certain operators, and in this paper as “Pressure Index (PI)”, h is defined as True Vertical Depth Sub Sea (TVDSS) with the unit in m,  P is pressure with unit in psi.

Salinity Affects Density

When salt is dissolved in fresh water, the density of the water increases because the mass of the water increases while volume remains the same. The more salt, which is dissolved in the water, the greater the salinity. When comparing two samples of water with the same volume, the water sample with higher salinity will have larger mass, and it will therefore be denser. These concepts are illustrated in Figure 2.

Figure 3 shows a direct relationship between salinity and water density, this relationship can be used to estimate water density from known water salinity with Equation as follows:

………………………………………………………….          (3)

We can also re-arrange Equation (3) in order to estimate water salinity from fluid density with the Equation as follows:

…………….………………………………………………………. (4)

From Equation (1) and Equation (3), we can get the following equations:

…………………………………………….          (5)

The above Equation is illustrated in Figure (4).

According to Howes (1998), the most commonly applied rule of thumb for normal hydrostatic gradients is 0.433 psi/ft (9.79 mPa/km). This is tied to the density of pure water at standard conditions. According to British Columbia Oil and Gas Commission (2014), standard conditions being 30 deg C and 101 kPa.

Actual normal hydrostatic gradients in formation are controlled by (a) brines of varying composition and density subject to PVT condition, in combination with (b) small scale pressure disequilibria (Howes, 1998).

The proposed equations (3 and 4) are a direct relationship between pressure gradient using PI and water salinities. These are simple equations which can easily be applied. Papers on the relationship between water density and salinity are previously discussed by Collins (1987), McCain (1991), and British Columbia Oil and Gas Commission (2014). These relationships from Collins and McCain are used by AAPG and SPE. However, AAPG’s Beaumont and Fiedler (1999) also suggested the use of a chart from Gearhart-Owens Industries (1972) to estimate formation water density from ppm NaCl and temperature. A similar approach is also introduced by McCutcheon, et. al. (1993), with water density as a function of temperature and salt concentration (salinity). The equations to estimate water density based on salinity only are tabulated in Table 2. Looking at the units and the equation format, Collins’s Equation is close to the PI equation.

Rogers and Pitzer (1982) give precise calculations, but very detailed. They tabled a large number of compressibility, expansiveness and specific volume values versus molality, temperature, and pressure. A semi-empirical equation of the same type was found to be effective in describing thermal properties of NaCl (0.1 to 5 molality) and was used to reproduce the volumetric data from 0 to 300°C and 1 to 1,000 bars.

The Relationship Between Pressure Gradient and Pressure Index

PI and Pressure Gradient have the same concept as they are a function of an over depth. At the same condition both pressure gradient and density equivalent will be the same no matter which TVDSS depth locations they are located.

According to Beaumont, et. al., (1999), the formation fluid pressure at any depth in a well is a function of the average formation water density above that depth, not the density of the formation water at any particular depth, formation water generally becomes denser with increasing depth

Hydrostatic pressure can also be calculated using following Equation:

……..………………………………………………….             (6)

By combining Equation (1) and (6) we can estimate density equivalent from pressure gradient with Equation as follows:

        …………..…………………………………………………………….         (7)

We can also re-arrange Equation (8) in order to estimate pressure gradient in psi/ft from density equivalent with the Equation as follows:

…………..……………………………………………………….       (8)

Application-1:  Pressure Comparison and Depletion Identification

Pressure Comparison

In Table 3, we can know that in reservoir X, the pressure of Well A is 2,000 psi at 1,400 mTVDSS and the pressure of Well B is 2,100 psi at 1,500 mTVDSS. Based on pressure information alone, it is hardly to define between the 2 wells which one we could classify as high or low pressure since both pressures were measured at different depths. By applying Equation (1), the PI of Well A is 1.004 g/cc and well B 0.984 g/cc. And referring to Figure 1, we could conclude that Well A has higher pressure than Well B.

Initial Pressure Prediction

In Table 2, reservoir X from well C has the pressure 2,000 psi acquired at depth 1,550 mTVDSS. From PI value if 0.907 g/cc this is showing an already depleted reservoir. For initial pressure we can estimate by using Equation (1), initial PI of 1.004 g/cc as calculated in well A. By applying initial PI and depth, we can estimate the initial pressure of reservoir X in Well C as 2213.8 psi.

Depletion Identification

From Table 4, there are two wells, VLM-2 is an exploration well, and VLM-A22 is a development well drilled post-production. All the pressure values are in psi unit, reservoir unit (sand) and depth in mTVDSS, pressure comparison from these two wells were challenging. For instance, Unit 1.2, in VLM-2 well the pressure was 1794 psi, and VLM-A22 was 1862 psi. This does not mean that VLM-A22 (1862 psi) has higher pressure than VLM-2 (1794 psi) as displayed in Table 5, VLM-A22 may have pressure a 68 psi higher than VLM-2 but the pressure data were acquired at different depth. By calculating the PI for the measured pressure at different depths, we observed that Unit 1.2 VLM-2 has PI 1.02 g/cc showing virgin pressure condition (PI value is more than 1 g/cc) while PI of VLM-A22 is 0.98 g/cc showing depleted pressure (PI value is less than 1 g/cc) as shown in Table 6.

Application-2: Pressure Gradient Prediction

Water resistivity is required first to calculate the water saturation, and the formation salinity can be estimated from resistivity using an empirical relationship.

From Figure 5, this shows that the water resistivity using Picket plot of Stage IVC Sand is 0.18 ohm-m. The selected interval for the depth 1,742-1,744 mTMD (1,677-1,679 mTVDSS), based on temperature log this sand has temperature 220 deg F. The estimated water salinity based on Rw of 0.18 ohm-m and FTEMP of 220 deg F is 13,000 ppm. These values are tabulated in Table 7.

As the salinity value is obtained, we can estimate the PI and pressure gradient over water interval. With salinity of 13,000 ppm, we can estimate the water density with several approaches, i.e. PI (as per equation 4 above), British Columbia Oil and Gas Commission (2014), McCain (1991), and Collins (1987) as we can see in Table 7. From this salinity of 13,000 ppm, we can estimate the water density with the above methods. The water density from PI is very close to McCain and Collins methods as tabulated in Table 8. In Table 9, we can see that pressure gradient from PI has similar value to the McCain and Collins approach. The pressure gradient is 0.437 psi/ft.

Application-3:  Formation Pressure Prediction

In Table 10, in Well-Y, where we have four pressure test measurements in oil-bearing reservoir A, B, C, and D, reservoir C has no pressure test value due to tight formation pressure measurement. To predict formation pressure in sand C, pressure test data from sand A, B, and D are available. Since PI is a function of hydrocarbon column thickness, the thicker the hydrocarbon column, and the higher the PI value.

To estimate the PI range at a certain depth we need to know the relationship between PI and depth in TVDSS. First, we have to calculate PI values from these three reservoirs resulting in PI ranges between 1.011 and 1.016 g/cc. The second stage is to establish the PI-TVDSS relationship. Figure 6 presents the relationship between depth in TVDSS and PI in g/cc, an empirical equation was obtained showing TVDSS= (-5.942.7*PI) +7502.1. By re-arranging this relationship as PI= (7502.1-TVDSS)/5942.7.

The third stage is to calculate the PI value at 1,487 mTVDSS which give the PI range at 1,487 mTVDSS as 1.012 g/cc, and we can predict the formation pressure using equation 1. The pressure at depth 1,487mTVDSS can be predicted as 2,140.3 psi. By validating the predicted one with a true measured pressure we find that, the average pressure difference is 0.7 psi and average percentage difference is 0.03 %. Hence it gives us a good confidence to use this approach to predict formation pressure where no formation pressure is measured.

Table 11 shows that the predicted pressure has a value close to measured formation pressure from wireline. The difference is very negligible between 0.04 and 3.73 psi or between 0.28 % and 0.46 %.

Application-4: Quick Look Fluid Typing

As we discussed earlier, PI can be used as a tool for quick look fluid typing. Table 12 showed the fluid typing based on PI method applied in VLS-1 well. Fluid types in column 6 are based on pressure plot and qualitative log interpretation, and fluid types in column 7 is based on quick look fluid typing using the PI approach, which shows a very high agreement between both approaches. The fluid typing is performed by examining the PI value differences, water has a PI of less than 1.012 g/cc, oil has PI 1.013-1.019 g/cc and gas has PI greater than 1.020 g/cc.

Some limitations of quick look fluid typing using PI were highlighted.  The PI cut-off should be used in the same pressure system. Different pressure system may result in different PI value ranges for each fluid type. In the water interval, PI value should be validated with salinity data information. Fluid type using PI can only be applied when reservoir pressure is in a hydrostatic condition. If it is in hydrodynamic conditions, the PI method can not be applied. Moreover, fluid typing in the transition area between fluid contact depths are the intervals with high uncertainty. However, this method presents an efficient way for early prediction of fluid type that can be used in reasonable time. The method is user friendly, straightforward and can therefore be applied to other fields.

This quick look fluid typing is an interim approach to predict fluid type with limited data information. Therefore, it is not considered a final fluid type. Appropriate fluid type determination should be based on integrated fluid typing and when the main information for fluid type becomes available, i.e. pressure-depth plot, fluid analyses, and DST.

This fluid typing using PI method was applied in three fields in Offshore Sabah, Malaysia and calibrated with 144 pressure test data acquired in seven wells. Although fluid typing using PI method is arbitrary and very interpretative, it shows 88% of fluid typing using PI method matches with fluid typing based on pressure plot analysis (Table 13). The seven wells are in initial condition, with no production during drilling. This method has not been tested in depleted and non-clastic reservoirs.

Conclusions

A vast characterization of reservoir pressure is needed for more focused decisions where the data is partially acquired. In order to unlock this value, an easy practical equation is required without advance computation and fancy graphical plots. PI can be calculated quickly and easily by PI= Pressure in psi/ (TVDSS depth in meter x 1.422).

The conventional pressure gradient unit is in psi/ft, kPa/m or ppg, while PI unit is in g/cc. This PI method is based on the concept from natural phenomenon of water density combined with pressure and depth as the fundamental relationship in the pore pressure characterization. At the same water characteristics, PI will be the same no matter which TVDSS depth locations they are located.

Far  more   than   a   simple   tool based on a plain equation, PI is a powerful decision analysis approach that adds   value   and   gives   insight   to   provide a baseline reference to determine and analyze pressure conditions as to whether overpressure, normal hydrostatic or depleted, without applying a conventional approach (i.e. graphical plots between pressure and depth, pressure gradient and depth).  In the PI concept, overpressure is referred as PI higher than Water PI (1.00 g/cc for salinity less than 2.5 kppm and 1.08 g/cc for salinity more than 300 kppm). This is also a perfect indication of hydrocarbon presence. While subnormal or depleted pressure is referred as lower PI than Water PI due to post-production.

The technique of PI not only benefits its users as a tool for screening reservoir pressure conditions, but can also help characterizing the temperature, water salinity, and water pressure gradients.

Other than that, PI can be applied as a quick look indicator to differentiate fluid types. Fluid type approach refinement is expected when the main information for fluid type is available. A holistic fluid typing should be conducted for a very robust fluid type determination. Initial pressure condition can also be predicted using PI when pressure data post-production is known.

The methodology proposed here is simple and fast, but robust enough to account for limited data and while full data acquisition is in progress (partially obtained). The method can help the subsurface team in understanding the pressure nature in quick turnaround time prior to the completion of ongoing data acquisition. Two best practices of this pressure evaluation using PI are very quick and robust. PI is very helpful for operations which require faster decisions and strong technical justification.

Acknowledgements

The authors are grateful to the colleagues in PETRONAS’s Petroleum Engineering Department for providing the constructive feedbacks and Mr. Zaki Sakdillah (Former Head of Reservoir Petrophysics) who provided his supports to publish this paper.

Nomenclatures

GRAD= Pressure gradient (psi/ft)

P = pressure (psi)

PI = Pressure Index (gram/cubic centimetre)

TVDSS= True Vertical Depth Sub Sea (m)

SAL= Salinity (ppm)

References

Beaumont and Fiedler, 1999, Treatise of Petroleum Geology: Formation Fluid Pressure and Its Application Formation Fluid Pressure and Its Application, 5-13.

Collins, A.G. 1975. Geochemistry of Oilfield Waters. New York: Elsevier Scientific Publishing Co.

Gearhart-Owens Industries, 1972, GO Log Interpretation Reference Data Handbook: Fort Worth, Gearhart-Owens Industries Inc., 226

Howes, J., 1999, WFT Data and Pressure Profile Analysis: Applications, Complexities, and Challenges, 27th Annual Convention Indonesian Petroleum Association, 1.

Howes, J., 1998, A Critical Evaluation of Normal Hydrostatic Gradients with an Example from Offhore SE Asia, ELF Bulletin des Centres de Recherche Workshop on Overpressures in Petroleum Exploration, 1-2.

McCain Jr., W.D. 1991. Reservoir-Fluid Property Correlations-State of the Art SPE Reservoir Engineering May 1991: 266-272. SPE-18571-PA.

McCutcheon, S.C., Martin, J.L, Barnwell, T.O. Jr. 1993. Water  Quality in Maidment,. Handbook of Hydrology, McGraw-Hill, New York,11.3

Mouchet, J.P. and Mitchell, A., 1989 Abnormal Pressures While Drilling: Origins, Prediction, Detection, Evaluation, Technip

Seraphin, et al., 2017: Density, Temperature, and Salinity, https://manoa.hawaii.edu/exploringourfluidearth/physical/density-effects/density-temperature-and-salinity, [access on 26 December 2017)

Figure 1: Formation Pressure Classification

Figure 2: The cubes and green coloured forms in this figure show the effects of varying mass and volume on density (Seraphin, et. al., 2017).

Figure 3: Relationship of Salinity and Pressure Index (PI)

Figure 4: Relationships Between Water Salinity and Hydrostatic Pressure

Figure 5: Water Resistivity of Stage IVC Sand in Zimba-2

Figure 6: PI-TVDSS Plot of Well-Y

Table 1: Tabulation of Salinity and Density from Water (Mouchet et. al, 1989)

Table 2: Some Approaches Used to Estimate Water Density

Table 3: Pressure Comparison of Reservoir X between Well A and Well B

Table 4: Pressure Comparison between VLM-2 and VLM-A22

Table 5: Pressure Comparison between VLM-2 and VLM-A22 with Additional Depth Information

Table 6: Pressure Comparison Using PI Method

Table 7: PI Prediction Based on Water Resistivity

Table 8: Comparison of Water Density Estimation Between PI and Other Approaches

Table 9: Comparing Pressure Gradient Prediction Between PI and Other Approaches

Table 10: Comparison between Predicted Pressure and Measured Pressure

Table 11: Comparison between Predicted Pressure and Measured Pressure of Well-Y

Table 12: Fluid Typing Using PI Method in VLS-1 Well

Table 13: Fluid Typing Using PI Method Applied to Three Fields