PORE-PRESSURE PLOT ACCURACY INCREASED BY MULTIPLE TREND LINES

May 7, 1990
Joe Foster Jr. M-1 Drilling Fluids Co. Houston A detailed, computer-assisted procedure has been developed for plotting pore pressures from electric logs. This new procedure differs from the traditional procedure in two ways. For atypical 10,000-ft well, 200-250 data points are plotted instead of the traditional 10-30 data points. The new procedure also uses multiple compaction trend lines instead of a single compaction trend line. This provides a separate trend line for each geologic age and
Joe Foster Jr.
M-1 Drilling Fluids Co.
Houston

A detailed, computer-assisted procedure has been developed for plotting pore pressures from electric logs. This new procedure differs from the traditional procedure in two ways.

For atypical 10,000-ft well, 200-250 data points are plotted instead of the traditional 10-30 data points.

The new procedure also uses multiple compaction trend lines instead of a single compaction trend line. This provides a separate trend line for each geologic age and fault sequence instead of one trend line throughout the entire plot.

This is important when plotting logs from wells penetrating more than one geologic age. Wells in deep water offshore where abnormal pressures are encountered in the Pleistocene and Pliocene as well as the Miocene are prime examples.

Faults, changes in geologic age, lithology, and salinity can be identified in the new procedure.

Adjustments can be made to offset the effects of these changes by shifting the trend lines. The correct positioning of the trend lines for each interval is self-revealing to the experienced plotter when enough data points are plotted.

The traditional procedure lacks sufficient detail to identify these changes.

To compensate, various methods are applied in the traditional procedure to manipulate the data before they are plotted.

TRADITIONAL METHOD

In a normally pressured environment, compaction of sediments below the earth's surface increases with depth due to overburden loading. As compaction increases, the porosity decreases.

The reduction in porosity results from a leak-off of the fluid content in the somewhat permeable sediments.

Because water in the pore space is a better conductor of electrical current than the rock matrix, the total measured conductivity will decrease with depth. This decrease in conductivity is seen as a straight line when plotted on semilog paper as illustrated in Fig. 1, where a decreasing trend is evident.

Because resistivity is the inverse function of conductivity, it will increase with depth in a normally pressured environment.

For overpressures to occur, a seal or section of low permeability (such as the geopressured shale itself) must be present to prevent formation fluids from migrating as overburden loading increases. Because fluids in the sediments below the seal are trapped, the formation below the seal will have more porosity than the formation just above the seal.

When log values (conductivity, resistivity, interval transit time, etc.) are plotted through normal pressures, a normal compaction trend is established. By plotting these values through a seal and into overpressures, the increase in porosity (increase in formation water) below the seal will cause a reversal of the trend, or deviation of the data points from the normal compaction trend.

Formation pore pressures are estimated in equivalent mud weight gradient by the magnitude of displacement of the data points from those along the normal compaction trend line. This is the basis for the most widely used method of pore pressure depiction from well logs.1-7

COMPACTION TRENDS

About 25 years ago, technology was developed to estimate formation pressures by relating log data to formation compaction.1 2 In this way, a normal compaction trend is established from a plot of shale conductivity/resistivity or interval transit time values which usually consist of 10-30 data points. Deviation from this normal compaction trend indicates an increase in pore pressures.

Overlays were developed to reflect changes in the average slope of the normal compaction trend by the geologic age of the formation.2 These overlays also reflect the pore pressure calibration by the spacing of the mud weight gradient lines for an age.

These were giant steps forward in drilling technology. It permitted construction of pressure profile maps which showed the pressure seal (the top of overpressures) and helped locate the important protective casing point.

The single trend line positioned through this seal provided a plot which indicated the top of overpressures. This is used to prevent drilling underbalanced into a major shale and reduce the probability of encountering a kick when the next sand is drilled. But this technique often exaggerated the magnitude of overpressures.

Because most of the early plotting was done on inland marsh wells of South Louisiana where Miocene sections are thick and relatively uniform, pore pressure depictions were reasonably dependable, since the depth intervals plotted were relatively short. But as this concept was expanded to include shallower and greater depths, and in other areas in which formations other than Miocene were present, its inconsistencies became obvious to those using it.

NORMALIZING PLOTS

Over the years, modifications were made in attempts to overcome these limitations. Some pore pressure plotting analysts devised methods to "normalize" log data points before plotting when these data points appeared to be influenced by changes in salinity or lithology.

These normalizing procedures include ratios or correction factors to be applied to data points when affected by increases in salinity and to limey shales. Little has been published on these procedures since the method often varies from one individual to another.

However, one normalizing procedure 3 which was published, involves applying the ratio of the short normal resistivity value at the point of interest and the short normal resistivity value from a depth of normal pressure to the resistivity or conductivity value to be plotted. Fig. 2 illustrates the type of plot resulting from these normalizing procedures.

However, correction factors are often applied so that the estimated pore pressures match the mud weight used to drill the well, instead of being obtained from the log data independently and interpreted in a consistent manner to estimate pore pressures. This highly subjective technique is still the most widely used plotting procedure today.

MULTIPLE TREND LINES

A detailed, computer-assisted procedure is now being successfully used to plot pore pressures. This multiple trend line procedure uses 250-300 uncorrected data points.

These data points are entered into the computer and plotted directly onto semilog paper. This eliminates the manipulation of data points.

Formation changes, faults, lithology, and other inconsistencies which cause an interruption in the normal compaction trend can be identified. Interruptions in the normal compaction trend result in a shift of the trend. After the shift, a new trend begins.

This shift in the normal compaction trend is illustrated in Figs. 3 and 4. Therefore, each plot will have several trend lines. This allows the proper overlay to be matched to individual trend sequences.

This is important because the slope of the normal compaction trend of each geologic age and the spacing between the mud-weight calibration lines differs significantly as expressed in the individual overlays. This type of plotting is illustrated in Figs. 5-8.

SELECTION OF OVERLAYS

If a formation (such as older, harder, or more calcareous rocks) is deposited over a relatively long period of time, the increase in formation compaction with depth is moderate.

This results in a moderate or more nearly vertical slope of the normal compaction trend line.

The more rapidly deposited, loosely consolidated marine sands and shales are compacted at a more rapid rate with increasing burial depth. This results in a greater slope of the normal compaction trend line from vertical (Fig. 9).

These compaction trend slope changes for each geologic age can be seen in the published work of Matthews and Kelly,2 which was based on many wells along the Louisiana and Texas Gulf Coast, and is still considered the most reliable method of estimating pore pressures from well logs.

Some who have plotted logs from many parts of the world are of the opinion that the slopes for the normal compaction trend lines, which Matthews and Kelly established for each geologic age, are repeated consistently throughout the world in their respective geologic ages.

OVERLAY SELECTION

When selecting the proper overlay, the following principles should be observed:

  • Each overlay should be used only in the geologic age for which it was intended (i.e., Miocene overlay used only in a Miocene age formation).

  • When the normal compaction trend shifts, the normal line on the overlay should be shifted accordingly.

Conductivity/resistivity, and interval transit time overlays have been developed from "Matthews and Kelly" and other technology for Pleistocene/Pliocene, Miocene, Frio (Oligocene), Vicksburg (Oligocene), Wilcox (Eocene), and older hard rock formations.

Each geologic age overlay not only has a different slope of the normal line (except for the Frio and Vicksburg, which are both in the Oligocene age, but different geology), but the spacing of the mud weight gradient calibration lines are also quite different. Therefore, selection of the proper overlay is very important.

There are several ways to identify the various geological ages and determine the proper overlay(s) to use. Some of these are listed below:

  • Knowledge of the area in which the well is being drilled, which includes the compaction trend sequence. Fig. 10 is the typical compaction trend sequence for the Gulf Coast area.

  • Geological markings such as formation tops, etc. on the log (if marked).

  • Obtaining the formation tops from the geologist.

  • The slope of the plotted log data being parallel to the normal compaction trend line of one of the overlays.

  • Relating paleo markings on the log (if marked) to the paleostratigraphic column for the area.

Caution: In deep water Gulf of Mexico environments where the upper Miocene consists mostly of rapidly buried sands, 8 fossils may not exist, because fossils are formed in shales. For example, the absence of Miocene fossils where they would normally be expected to exist, may lead to the erroneous assumption that the sediments are of the Pliocene geologic age.

Therefore, if the top of the Miocene is picked from paleo data alone, this could result in picking the top of the Miocene much lower than it actually is. This makes tremendous difference in pore pressure interpretation.

For this reason, comparing the compaction slope of the Miocene normal line on the overlay to the compaction slope of the data points on the plot is often more helpful in picking the top of the Miocene than using paleo data.

  • Five general shale types are:

    1. Pliocene-Alternating sand/shale sequences, but mostly sand. Resistivity is usually greater than 1.0 ohmmeter.

    2. Miocene-Alternating sand/shale sequences with resistivity less than 1.0 ohmmeter.

    3. Frio-High resistivity, limey shales.

    4. Vicksburg-Low resistivity, steeply compacting argillaceous shales.

    5. Wilcox or hard rock-Sandy, very high resistivity (usually greater than 5.0 ohm-meter), low rate of compaction shales.

  • Trial and error fit with mud weight used (i.e., a Miocene overlay placed over widespread Pliocene data points would indicate high pore pressures, when the interval may have been drilled with a 9.0-9.5 ppg mud weight. On the other hand, a Pliocene overlay placed over more tightly spaced Miocene data points would indicate very low pore pressures when a 12.0 ppg mud may have been required to drill the interval).

  • Known pressure points such as repeat formation tests (RFT's), drill stem tests (DST's), kicks, production and reservoir data, etc.

    Caution: Overreacting to gas-cut mud is a common practice when drilling below intermediate casing. Gas-cut mud by itself is a very poor indicator of pore pressures. The practice of increasing the mud weight to overcome gas-cut mud often exposes the formation to induced pressures far greater than confirmed formation pore pressures.

    Production and reservoir data are the most reliable method of determining formation pore pressures. If these data are not available, formation pressures calculated from shut-in drill pipe pressure readings from a kick should be considered more reliable than low volume pressure tests such as RFT's.

    RFT's taken in sands exposed to induced pressures usually reflect the induced pressure instead of the formation pressure. When this occurs, the RFT will reflect pressures equivalent to, or near, the mud weight being used.

  • Limey caps or major sand bodies help pick positions of compaction trend lines and separate shifts from one geologic age or shale from another.

OVERLAY USE

Conductivity/resistivity overlays are designed to interpret conductivity plots with the printed side face up. Because resistivity is the inverse function of conductivity (1/1000), the conductivity overlay should be turned face down to interpret resistivity plots. Interval transit time overlays should be used with the printed side up.

  • The properly selected overlay should be used in a precisely vertical position relative to the plot. This can be achieved by aligning the vertical margin of the overlay with any vertical line on the plot paper. This keeps the slope of the normal line on the overlay at the proper angle, which provides a more accurate plot and also helps differentiate between various geologic ages.

    Exceptions to this are directional wells in which the vertical border is tilted as marked on the overlay to correspond with hole deviation from vertical.

  • When selecting the placement for the normal compaction trend line, slide the overlay to the left or right over the plot, maintaining its vertical orientation until the normal compaction trend line passes through the most likely compaction trend for a given interval.

    Look for small zig-zag inflection points, in near-equilibrium with each other, which follow a straight trend line of the proper compaction slope. In placing the trend line, avoid the large excursions on the plot. The more of these short, slightly different data points plotted, the more accurate the trend line placement will be.

    Look for shale data points Immediately adjacent to, or stringers within, permeable sand bodies. These points often approach the normal trend line. Avoid extremely high or low values caused by salt water or hydrocarbons within the adjacent sands.

  • Look for thick, limey shale caps which extend far beyond the normal trend line. These caps often indicate major pore pressure increases or decreases, fractures, faults, or other inconsistencies which could indicate a trend line shift; or possibly a casing point.

    Also look for large zig-zag inflection points which exceed the normal range of data points for a specific interval. This is also a possible indicator of a fault or geologic age change, especially if another trend is apparent below the zig-zag.

  • Expect several short trend lines at shallow depths which shift to the left or right as different Pleistocene, Pliocene, and Miocene depositional blocks (probably due to the different water depths in the Gulf during the ice ages) are penetrated.

Note that as the normal compaction trends shift, the slope remains the same as long as the geologic age remains the same.

In deep water off the Continental Shelf where the bedding planes slope steeply downward, shale resistivities drop to very low values (0.10.2 ohm-meters).

Therefore, the trend lines must be shifted accordingly, or the pore pressures will be greatly over estimated (Fig. 8).

Because plotting pore pressures from well logs is not a direct measurement of pore pressures, the interpretation should always be subject to verification and/or modification.

Compare the plot to the mud weights used to drill the well. These mud weights are not always in agreement with the actual formation pressure.

A significant disagreement should be accompanied by an explanation of the difference, when the data are available.

Experience shows that most operators drill nearly balanced to 1,000 ft below intermediate casing, but overbalanced below that point. This overbalance is usually the result of overreacting to drill-gas expansion at the surface.

This gas can come from a sandy interval which has a pore pressure no greater than, and in some cases less than, the previous shale sections drilled.

Therefore, in these cases the increase in mud weight is not needed.

The plot can reveal reasons for drilling and hole problems when such problems were caused by pore pressure changes which were not immediately balanced by mud weight changes.

These problems may include drilling breaks, kicks, connection gas, sloughing shale, hole enlargement, tight hole, bridges, fill, lost circulation, and differentially stuck pipe.

INCREASED ACCURACY

Plotting pore pressures from well logs is not a direct measurement of pore pressures, but an interpretation of pore pressures based on normal shale compaction vs. undercompaction. It is, at best, technology that includes considerable freedom of interpretation. However, this detailed, computer-assisted procedure places detailed, uncorrected data on the plot. Therefore, all interpretation is done on the plot where it can easily be scrutinized.

Furthermore, each overlay is used only for its proper geologic age/lithology compaction characteristics-not one overlay and one trend line for the entire plot.

To produce this type of plot, greater attention to detail regarding the formation age, lithology, and other variables is required. The higher degree of integrity in this procedure has resulted in a more accurate pore-pressure prediction.

ACKNOWLEDGMENT

I would like to thank the management of M-1 Drilling Fluids Co. for allowing me to write this article. The cooperation of personnel at the Houston technical center is appreciated. Credits are given to James Gill, Jim Bruton, and Jay Bruton who contributed a great deal to this article.

REFERENCES

  1. Hoftman, C.E., and Johnson, R.K., "Estimation of Formation Pressures from Log-Derived Shale Properties," Journal of Petroleum Technology, June 1965, pp. 717-22.

  2. Matthews, W.R., and Kelly, J.. "How to predict formation pressures and fracture gradient," OGJ, Feb. 20, 1967, pp. 92-106.

  3. Greene, Ken, "Normalizing technique helps plot pressures from logs," OGJ, Oct. 23, 1978, pp. 147-56.

  4. Gill, J.A., "Hard Rock Drilling Problems Explained by Hard Rock Pressure Plots," IADE/SPE paper 11377, Drilling Conference, New Orleans. February 1983.

  5. Gill, J.A., and Gregg, D.N., "Locating the Dog-Legs and Jump-Shifts in North Sea Pressure Profiles," SPE paper 5262, London, April 1975.

  6. Gill. J.A., "The Tuscaloosa-Woodbine Trend: Tracking Its Unique Pressure Profiles," SPE paper 10235, San Antonio. October 1981.

  7. Gill, J.A., (pore pressure consulting engineer), personal communication, 1989.

  8. Dunlap, Joe, (paleontologist, retired), personal communication, 1989).

Copyright 1990 Oil & Gas Journal. All Rights Reserved.