Skip to Main Content

Abstract

This chapter presents two case studies where resistivity-tool-response modeling was applied for formation evaluation. Induction logs from one field were modeled for a number of deviated wells assuming that mud-filtrate invasion was negligible (i.e., by applying 1-D modeling). In the second example, laterolog (DLL) logs from another field show clear separation between the deep and shallow curves, indicating possible invasion. Therefore, 2-D modeling was applied in this case.

These two case studies are not meant to comprise an exhaustive set of examples for the application of resistivity modeling. They are simply the examples that were available for release at the time of writing.

Overview

Introduction

This chapter presents two case studies where resistivity-tool-response modeling was applied for formation evaluation. Induction logs from one field were modeled for a number of deviated wells assuming that mud-filtrate invasion was negligible (i.e., by applying 1-D modeling). In the second example, laterolog (DLL) logs from another field show clear separation between the deep and shallow curves, indicating possible invasion. Therefore, 2-D modeling was applied in this case.

These two case studies are not meant to comprise an exhaustive set of examples for the application of resistivity modeling. They are simply the examples that were available for release at the time of writing.

ILD 1-D Modeling Example: Deviation/Dip Effect

Case study 1: Background

The Prudhoe Bay oil field of North America began production with estimated in-place reserves of about 22 billion barrels of oil and 47 TCF of gas (McCoy, 1997). The field is a combination structural-unconformity trap with hydrocarbons reservoired in fluvio-deltaic sandstones of the Permian–Triassic Ivishak Formation. To date, more than 1500 wells have been drilled in the field. Because the drilling is conducted from spaced gravel pads, a majority of the wells have wellbore deviations in excess of 20°–30°. Neither the response of deep induction curves in these deviated wells nor the ultimate impact on log-derived Sw and HPV is well known. In this case study, 1-D modeling was performed to correct for deviation and bed-thickness effects on ILD logs in 16 wells. The potential impact on Sw and HPV estimates was also studied.

Wells with deviations greater than 30° were selected for Phase 1. Ten wells met this criterion. A second phase focused on several wells for which oil-based core (OBC) existed, and the water saturation from the OBC was the standard used for reserve estimation. Ten wells also met this criterion, of which four had deviations greater than 30°.

Benchmark

Wellbore deviation causes horns or spikes around bed boundaries in induction logs as shown in Figure 3.3. This effect has been recognized since the mid-1980s, and technology to quantitatively compute the dip effects on the response of the industry-standard dual induction tool was developed by Hardman and Shen (1986). Assuming negligible borehole and invasion effects, the boundary-value problem of a magnetic dipole and its resulting electromagnetic field is solved for the induction sonde passing through beds dipping with respect to the borehole.

To address the impact of modeling the dip effect on ILD response in Prudhoe Bay wells, a preliminary modeling benchmark study was conducted. In the benchmark study, the center-bed thickness was varied from 5 to 100 ft (1.5 to 30.5 m) for both 0° dip and 45° dip. The center-bed resistivity was assumed to be 40 ohm m, and the adjacent shoulder beds were assigned 1 ohm m resistivity.

Figure 5.1 shows the ILD response for beds with differing apparent thickness in measured depth (MD). For the case of a horizontal 5 ft (1.5 m) (MD) bed, the computed induction log, ILD_C, reads about 63% of Rt, or 25 ohm m versus 40 ohm m, as a result of shoulder-bed effects. When the apparent bed dip is increased to 45°, the ILD_C in the 5 ft (1.5 m) (MD) bed is only 20% of Rt: 8 ohm m versus 40 ohm m, primarily because the true-vertical-depth thickness of this bed is now only 3.54 ft (1.08 m).

Figure 5.1.

Results from the benchmark study for beds dipping at 0° and 45° for variable bed thickness in measured depth (MD). The beds (dotted rectangular outlines) vary from 100 to 5 ft (30.5 to 1.5 m). Formation resistivity was 40 ohm m, and shoulder beds resistivity was 1 ohm m. The results show that dip effect on ILD response is nonlinear as a function of bed thickness. 1 ft = 0.3 m.

Figure 5.1.

Results from the benchmark study for beds dipping at 0° and 45° for variable bed thickness in measured depth (MD). The beds (dotted rectangular outlines) vary from 100 to 5 ft (30.5 to 1.5 m). Formation resistivity was 40 ohm m, and shoulder beds resistivity was 1 ohm m. The results show that dip effect on ILD response is nonlinear as a function of bed thickness. 1 ft = 0.3 m.

Figure 5.2 shows the results of the same benchmark study for true bed thickness (i.e., true vertical depth, TVD). For the dipping beds, the apparent bed thickness (MD) is thicker than the true vertical bed thickness by a factor of the cosine of the deviation angle. The reduction in resistivity in the dipping beds is similar but not identical to that observed in Figure 5.1 because of the differences in apparent bed thickness. The differences between the red and green bars reflect the true dip effect.

Figure 5.2.

Results from the benchmark study for beds dipping at 0° and 45°, compared in true bed thickness (TVD). Note that the TVD bed thickness is thinner than MD bed thickness. The reduction in resistivity in the dipping beds is similar but not identical to that seen in Figure 5.1 because of the differences in apparent bed thickness. 1 ft = 0.3 m.

Figure 5.2.

Results from the benchmark study for beds dipping at 0° and 45°, compared in true bed thickness (TVD). Note that the TVD bed thickness is thinner than MD bed thickness. The reduction in resistivity in the dipping beds is similar but not identical to that seen in Figure 5.1 because of the differences in apparent bed thickness. 1 ft = 0.3 m.

The effect of dip on ILD response is a nonlinear function of bed thickness and resistivity contrast between the formation and shoulder bed. Because service company chart books illustrate only specific, limited sets of assumptions, the actual impact on ILD response from the combined effects of variations in dip and bed thickness can only be addressed through tool-response modeling.

Bed boundaries

Bed boundaries are a required part of the earth model for the resistivity-forward-modeling process. For this case study, bed boundaries were defined based on the petrophysical response of logging tools with higher vertical resolution than the deep induction tool. Modeling software was used to square either the MSFL or SFL curve for wells with water-base mud, whereas for oil-based mud wells, the sonic (At) curve or bulk density (RHOB) curves were used. These curves have comparable vertical resolution (1–2 ft [0.3–0.6 m]) and have a few times higher resolution than the ILD (5–7 ft [1.5–2.1 m]). The derived petrophysical bed boundaries were then visually cross-checked with other lithological or geological bed boundaries defined by GR, RHOB, and neutron porosity (NPHI) logs to ensure consistency.

Forward-modeling-based iterative inversion

After the bed boundaries were derived, each bed was assigned an initial Rt resistivity based on either the maximum or minimum ILD value in the corresponding bed. Next, an automatic ILD inversion procedure was applied to find a final Rt value for each bed. This procedure works by computing the ILD response (ILD_C) for the given bed boundaries and resistivity values. It then adjusts the Rt values for each bed to minimize the differences between the field ILD log and computed ILD_C using an iterative least-squares minimization approach. The number of iterations was set to five, which resulted in a relative error between ILD and ILD_C of less than 3% in most cases. In some extremely high- and low-resistivity-contrast zones (e.g., around the thin pyrite intervals found in the Ivishak formation), a further iterative and time-consuming adjustment of the Rt values and bed boundaries was performed to minimize the difference between ILD and ILD_C.

Sensitivity to well deviation

To study the sensitivity of the ILD response to the deviation angle, the deviation for one well (D-18) was varied from the given 49.3° by five degrees (49.3±5), and ILD_C was also computed for zero deviation. The results are plotted in Figure 5.3 and show that a five-degree change in deviation angle did not significantly change the ILD response. In a bed with an apparent thickness of 10 ft (3.0 m) (at 11,910 ft in Figure 5.3) the actual deviated-well response is about 40% reduced from the vertical-well response. Most importantly, the ILD response is 7 times less than Rt in the same 10-ft (3-m) bed, illustrating the severity of the shoulder bed effect in this case where the shoulder-bed resistivity is about 7 ohm m.

Figure 5.3.

Depth plot for a well from the Rt inverse modeling used for the sensitivity study on dips. The dip was varied from the given 49.3°, using 49.3° by 5°, and 0° dip was also modeled. 1 ft = 0.3 m.

Figure 5.3.

Depth plot for a well from the Rt inverse modeling used for the sensitivity study on dips. The dip was varied from the given 49.3°, using 49.3° by 5°, and 0° dip was also modeled. 1 ft = 0.3 m.

Potential impact on hydrocarbon pore volume

The differences in hydrocarbon pore-thickness (HPT) calculated from the modeled resistivity (taken as Rt) versus HPT calculated directly from the deep induction curve (ILD) were computed for each of the 15 modeled wells. The results are plotted in Figure 5.4. Not surprisingly, the largest difference in HPT is associated with zones having smaller HPT values.

Figure 5.4.

Plot of percent difference in hydrocarbon pore thickness (a proxy for hydrocarbon pore volume) by zone as a function of HPT from deep induction log. 1 ft = 0.3 m.

Figure 5.4.

Plot of percent difference in hydrocarbon pore thickness (a proxy for hydrocarbon pore volume) by zone as a function of HPT from deep induction log. 1 ft = 0.3 m.

For the 15 wells, the HPT based on the modeled ILD curve increased an average of about 3% as compared to the HPT derived from the field ILD curve. The wells covered a range of wellbore deviations, from 0.5° to 49.3°. No obvious trend is observed between wellbore deviation and difference in HPT.

The zones with the biggest percent differences are zones 4B, 1A, and 1B. For zones 1A and 1B the differences can be explained by the presence of thin beds. For zone 4B, the differences result from significant shoulder-bed effects from the overlying shale.

Conclusions

Many apparently thick beds, as defined by GR and ILD alone, may be composed of thinner petrophysical beds that can be derived from SFL, MSFL, Δt, and RHOB logs.

In beds with apparent thickness less than 10 ft (3.0 m), the differences between the inverted Rt and measured ILD are significant, and modeling should be done to correct the bed-thickness and shoulder-bed effects even in the wells with no dip or with borehole deviations less than 30°. Dip alone is significant only in wells with an apparent deviation angle greater than 30°.

Differences between Rt and ILD logs are nonlinear. The ILD tool is sensitive to conductivity rather than to resistivity. Thus, beds with high resistivity values exhibit larger differences (in ohm m) between the Rt and the ILD than beds with low resistivity values. As shown in Figure 5.3, the ILD log responses are very close to Rt (the shaded squared red curve) in low-resistivity (high-conductivity) beds (left side of track) and far apart from Rt in high-resistivity (low conductivity) beds (right side of track).

A sensitivity study showed that a small variation in dip (e.g., ± 5 degrees dip uncertainty) around the wellbore deviation has no significant impact on the modeling results. This result is also applicable to boreholes with no deviation. Thus, small uncertainties in relative dip (the combination of borehole deviation and bed-dip angles) may not change inversion results significantly in this case study field.

In most beds, the measured apparent resistivity Ra (the field ILD log) does not agree with the inverted Rt data derived from modeling, which may be closer to the true formation-resistivity profile.

The inverted Rt data may not drastically change the saturation or HPV calculation compared with using ILD data in major, thick, producing zones. This is because the water-saturation calculation is no longer very sensitive to resistivity changes when the resistivity is higher than 100 ohm m.

DLL 2-D Modeling Example: Invasion Effect

Case study 2: Background

The wireline dual laterolog (DLL) resistivity and the logging-while-drilling (LWD) ring resistivity from an exploration well appear to be significantly different around a 1-m (3.3 ft)-thick formation that shows clear RHOB and NPHI crossover, indicating a potential gas-reservoir formation. The LWD ring resistivity peak reads about 20 ohm m, whereas the wireline dual laterolog deep (LLD) only measures around 5 ohm m at the same depth X54 (Figure 5.5). The decreased LLD reading at the base of the reservoir interval (around X320 m, Figure 5.6) may be a combination of thin-bed and shoulder-bed effects caused by the underlying shale bed. Separation in resistivity curves from a triple-combo (gamma ray, resistivity, and neutron-density tools) run in this well suggests that the porous and permeable reservoir units were invaded by mud filtrate.

Figure 5.5.

Composite depth plot of the thin-bed interval. In track 2, note the crossover between NPHI and RHOB around depth XX54. In track 3, R Ring from LWD shows around 20 ohm m; LLD reads around 5 ohm m. Track 4 shows the 2-D modeling result. Track 5 shows the potential impact of Rt modeling in Sw estimation. 2 m = 6.6 ft; RTSQ = square-shaped true resistivity given in a model; R Ring = ring resistivity; R BIT = resistivity measured at the drill bit.

Figure 5.5.

Composite depth plot of the thin-bed interval. In track 2, note the crossover between NPHI and RHOB around depth XX54. In track 3, R Ring from LWD shows around 20 ohm m; LLD reads around 5 ohm m. Track 4 shows the 2-D modeling result. Track 5 shows the potential impact of Rt modeling in Sw estimation. 2 m = 6.6 ft; RTSQ = square-shaped true resistivity given in a model; R Ring = ring resistivity; R BIT = resistivity measured at the drill bit.

Feasibility with tornado chart

Initially, the DLL tornado chart from the service-company chart book, plus a DLL environmental correction from 1-D radial-modeling results, were used for the invasion correction. This correction had negligible impact on the LLD, because the environment defined in the chart was not applicable for the downhole conditions of borehole diameter, bed thickness, and Rxo/Rm ratio.

Resistivity tool response modeling was carried out to understand the differences between the LWD and wireline resistivities and to obtain realistic water saturation estimates from resistivity.

Benchmark

A benchmark study was conducted first to investigate the impact of 2-D modeling based on the MSFL, LLS, and LLD logs. In 2-D modeling, resistivity can vary in both radial and axial directions in the borehole environment.

Figure 5.7 shows a 2-D forward-modeling code run over two hypothetical beds whose characteristics were derived from the study well, specifically from the bed shown at XX54 meters in Figure 5.5. Both beds were 3 ft (0.9 m) in thickness and had the same shoulder-bed resistivity, Rs = 1.5 ohm m, and the same formation resistivity, Rt = 20 ohm m. One bed was given an invasion radius of 15 in. (38 cm) and the other bed had no invasion. For the invaded bed, Rxo was set to 1.2 ohm m. Mud resistivity was Rm = 0.092 ohm m at 38°C and borehole diameter Db = 12.5 in. (32 cm). The results, plotted in Figure 5.7, show that LLD reads only 2.5 ohm m in the invaded bed, whereas the maximum LLD reading is close to 20 ohm m in the bed without invasion.

Figure 5.6.

Composite depth plot of the major reservoir interval modeled. Average water saturation decreased around 10-15% in the interval shown by using the modeled resistivity profile in place of LLD. This produced a 28% increase in gas-in-place for this major gas-reservoir interval of the well. 2 m = 6.6 ft; PHIT1, PHIT2 = total porosity from buttons 1 and 2; Vshale = volume of shale (%).

Figure 5.6.

Composite depth plot of the major reservoir interval modeled. Average water saturation decreased around 10-15% in the interval shown by using the modeled resistivity profile in place of LLD. This produced a 28% increase in gas-in-place for this major gas-reservoir interval of the well. 2 m = 6.6 ft; PHIT1, PHIT2 = total porosity from buttons 1 and 2; Vshale = volume of shale (%).

Figure 5.7.

Benchmark 2-D modeling over two hypothetical beds with characteristics derived from the study well. The lower bed is uninvaded and the upper bed has a 15 in. (38 cm) radius of invasion. 10 ft = 3.0 m; 1 in. = 2.5 cm.

Figure 5.7.

Benchmark 2-D modeling over two hypothetical beds with characteristics derived from the study well. The lower bed is uninvaded and the upper bed has a 15 in. (38 cm) radius of invasion. 10 ft = 3.0 m; 1 in. = 2.5 cm.

From the benchmark study, it was concluded that LWD resistivity likely reads true resistivity, whereas the DLL wireline log may be severely affected by salty mud invasion, and modeling DLL responses may result in a more realistic resistivity profile if the modeling is performed by considering the low-resistivity mud invasion, large borehole, and washout.

As a result of the findings from the benchmark test, 2-D modeling was run on the entire reservoir interval of the study well (Figure 5.6).

Formation microimage scaling and bed boundaries

As input to the resistivity model, fullbore formation microimager (FMI) data were used to produce As input to the resistivity model, fullbore formation microimager (FMI) data were used to produce produce FMI-based bed boundaries and a first-guess high-resolution apparent resistivity that is comparable in magnitude to the one from the LLS. The FMI-derived bed boundaries are verified and made consistent with lithological beds defined by the GR, NPHI, and RHOB logs by a manual editing process before the inversion begins.

Forward-modeling-based iterative inversion

To estimate Rt, a 2-D forward-modeling code for the dual laterolog tool was applied with an iterative inversion procedure. The forward-modeling code was executed repeatedly with adjustment of the formation resistivity and invasion radius in each iteration, until the agreement between the computed LLD_C and LLS_C with the field LLD and LLS logs showed less than 5% relative error. The whole interval took six iterations to converge to 5% relative error.

The iterative inversion process was monitored using the log display shown in Figure 5.8. The display has three tracks, all showing the bed boundaries that were established through an editing process that preceded the inversion itself.

Figure 5.8.

The depth plot displays a 2-D modeling work process. Logs in the left track provide lithological references; the 2-D borehole model in the middle track shows the borehole radius (RB) and invasion radius (RI); and the right track shows field resistivity logs (SRES, MSFL, LLS, and LLD). The right track also contains computed DLL logs under the given model Rt (RTSQ). As shown in the right track, the computed logs agree closely with the field LLD and LLS logs after inversion is completed. 10 m = 32.8 ft; 1 in. = 2.5 cm.

Figure 5.8.

The depth plot displays a 2-D modeling work process. Logs in the left track provide lithological references; the 2-D borehole model in the middle track shows the borehole radius (RB) and invasion radius (RI); and the right track shows field resistivity logs (SRES, MSFL, LLS, and LLD). The right track also contains computed DLL logs under the given model Rt (RTSQ). As shown in the right track, the computed logs agree closely with the field LLD and LLS logs after inversion is completed. 10 m = 32.8 ft; 1 in. = 2.5 cm.

The GR, NPHI, and RHOB logs are presented in the left track to provide lithological references. The middle track represents the 2-D borehole and invasion model. The mud resistivity Rm, the formation temperature, and the formation-water resistivity Rw have been entered in the modeling application’s parameter menu. The borehole radius (RB) is derived using the caliper log, and the invasion radius for each bed is iteratively adjusted by the user. The right track shows SRES, MSFL, LLS, and LLD curves. Rxo is initially defined by squaring MSFL curves. Rxo can be constrained by selecting minimum and maximum values in zones with constant caliper and GR. The Rxo value can be modified manually for each bed. The right track also contains the modeled Rt (RTSQ) and the computed LLD_C and LLS_C. The field DLL logs are plotted for comparison. Note that, at the end of the inversion, the computed LLD_C and LLS_C agree closely with the field LLD and LLS logs.

Inverted Rt and its impact on Sw estimation

For the main reservoir interval shown in Figure 5.8, the apparent LLD resistivity reads 10–20 ohm m between 270–300 m, whereas Rt derived from the 2-D inversion (shown as RTSQ in Fig. 5.8) approaches 40 ohm m. The reduction in LLD as compared to Rt is caused by salty mud invasion (Rm = 0.092 ohm m at 38°C).

Using the modeled resistivity profile, RTSQ, rather than LLD, resulted in a 15% increase in gas saturation for the interval shown, and a 28% increase for gas-in-place in this major gas-reservoir interval of the well (Figure 5.6).

The gas-saturation increase is even more significant for the thinly bedded upper interval, where the LLD resistivity was significantly lower than the LWD resistivity, even though there was a clear RHOB and NPHI crossover over the 91-cm (3-ft.) thick formation (XX54, Figure 5.5). The modeled RTSQ curve is plotted in track 5 of Figure 5.5. Around XX54, Rt is about 40 ohm m, where the LWD reads around 20 ohm m, and the DLL only showed around 5 ohm m. In track 6, SW(LLD) is the water saturation calculated with LLD and is the water saturation calculated from RTSQ. In the formation at XX54, SW(LLD) is about 50%, whereas SW(RTSQ) is only about 20%.

It is clear that 2-D resistivity modeling provides a means to estimate and correct for borehole and invasion effects on apparent DLL resistivity measurements, and to enhance vertical resolution of the reservoir units. This type of modeling, involving the integration of bed boundaries from FMI image analysis, provides a framework for improved estimates of Rt and, thus, log-based water saturation in such reservoirs.

Conclusions

Apparent resistivity as estimated from field logs (Ra) is not equal to true formation resistivity (Rt) because of the limitations in tool vertical resolution, depth of investigation, sensitivity, linearity, and dynamic range in a complex borehole environment with salty mud invasion.

Traditional chart book corrections do not remedy these complicated environments, but computer modeling can convert the apparent resistivity measured by logs into a response profile that may more closely resemble true downhole conditions.

To derive an accurate resistivity profile and reduce uncertainty, a high-resolution earth model defined by FMI images, cores, core description, EPT, high-resolution dipmeter, or other high-resolution logs is necessary.

Modeling Rt using geological and geophysical constraints appears to be appropriate and can be used for a more accurate and detailed estimation of hydrocarbon saturation.

With the development of log-modeling software that incorporates tool physics, computer modeling will replace conventional interpretation charts, and the log analyst will be able to interpret resistivity logs with increased confidence.

Figures & Tables

Figure 5.1.

Results from the benchmark study for beds dipping at 0° and 45° for variable bed thickness in measured depth (MD). The beds (dotted rectangular outlines) vary from 100 to 5 ft (30.5 to 1.5 m). Formation resistivity was 40 ohm m, and shoulder beds resistivity was 1 ohm m. The results show that dip effect on ILD response is nonlinear as a function of bed thickness. 1 ft = 0.3 m.

Figure 5.1.

Results from the benchmark study for beds dipping at 0° and 45° for variable bed thickness in measured depth (MD). The beds (dotted rectangular outlines) vary from 100 to 5 ft (30.5 to 1.5 m). Formation resistivity was 40 ohm m, and shoulder beds resistivity was 1 ohm m. The results show that dip effect on ILD response is nonlinear as a function of bed thickness. 1 ft = 0.3 m.

Figure 5.2.

Results from the benchmark study for beds dipping at 0° and 45°, compared in true bed thickness (TVD). Note that the TVD bed thickness is thinner than MD bed thickness. The reduction in resistivity in the dipping beds is similar but not identical to that seen in Figure 5.1 because of the differences in apparent bed thickness. 1 ft = 0.3 m.

Figure 5.2.

Results from the benchmark study for beds dipping at 0° and 45°, compared in true bed thickness (TVD). Note that the TVD bed thickness is thinner than MD bed thickness. The reduction in resistivity in the dipping beds is similar but not identical to that seen in Figure 5.1 because of the differences in apparent bed thickness. 1 ft = 0.3 m.

Figure 5.3.

Depth plot for a well from the Rt inverse modeling used for the sensitivity study on dips. The dip was varied from the given 49.3°, using 49.3° by 5°, and 0° dip was also modeled. 1 ft = 0.3 m.

Figure 5.3.

Depth plot for a well from the Rt inverse modeling used for the sensitivity study on dips. The dip was varied from the given 49.3°, using 49.3° by 5°, and 0° dip was also modeled. 1 ft = 0.3 m.

Figure 5.4.

Plot of percent difference in hydrocarbon pore thickness (a proxy for hydrocarbon pore volume) by zone as a function of HPT from deep induction log. 1 ft = 0.3 m.

Figure 5.4.

Plot of percent difference in hydrocarbon pore thickness (a proxy for hydrocarbon pore volume) by zone as a function of HPT from deep induction log. 1 ft = 0.3 m.

Figure 5.5.

Composite depth plot of the thin-bed interval. In track 2, note the crossover between NPHI and RHOB around depth XX54. In track 3, R Ring from LWD shows around 20 ohm m; LLD reads around 5 ohm m. Track 4 shows the 2-D modeling result. Track 5 shows the potential impact of Rt modeling in Sw estimation. 2 m = 6.6 ft; RTSQ = square-shaped true resistivity given in a model; R Ring = ring resistivity; R BIT = resistivity measured at the drill bit.

Figure 5.5.

Composite depth plot of the thin-bed interval. In track 2, note the crossover between NPHI and RHOB around depth XX54. In track 3, R Ring from LWD shows around 20 ohm m; LLD reads around 5 ohm m. Track 4 shows the 2-D modeling result. Track 5 shows the potential impact of Rt modeling in Sw estimation. 2 m = 6.6 ft; RTSQ = square-shaped true resistivity given in a model; R Ring = ring resistivity; R BIT = resistivity measured at the drill bit.

Figure 5.6.

Composite depth plot of the major reservoir interval modeled. Average water saturation decreased around 10-15% in the interval shown by using the modeled resistivity profile in place of LLD. This produced a 28% increase in gas-in-place for this major gas-reservoir interval of the well. 2 m = 6.6 ft; PHIT1, PHIT2 = total porosity from buttons 1 and 2; Vshale = volume of shale (%).

Figure 5.6.

Composite depth plot of the major reservoir interval modeled. Average water saturation decreased around 10-15% in the interval shown by using the modeled resistivity profile in place of LLD. This produced a 28% increase in gas-in-place for this major gas-reservoir interval of the well. 2 m = 6.6 ft; PHIT1, PHIT2 = total porosity from buttons 1 and 2; Vshale = volume of shale (%).

Figure 5.7.

Benchmark 2-D modeling over two hypothetical beds with characteristics derived from the study well. The lower bed is uninvaded and the upper bed has a 15 in. (38 cm) radius of invasion. 10 ft = 3.0 m; 1 in. = 2.5 cm.

Figure 5.7.

Benchmark 2-D modeling over two hypothetical beds with characteristics derived from the study well. The lower bed is uninvaded and the upper bed has a 15 in. (38 cm) radius of invasion. 10 ft = 3.0 m; 1 in. = 2.5 cm.

Figure 5.8.

The depth plot displays a 2-D modeling work process. Logs in the left track provide lithological references; the 2-D borehole model in the middle track shows the borehole radius (RB) and invasion radius (RI); and the right track shows field resistivity logs (SRES, MSFL, LLS, and LLD). The right track also contains computed DLL logs under the given model Rt (RTSQ). As shown in the right track, the computed logs agree closely with the field LLD and LLS logs after inversion is completed. 10 m = 32.8 ft; 1 in. = 2.5 cm.

Figure 5.8.

The depth plot displays a 2-D modeling work process. Logs in the left track provide lithological references; the 2-D borehole model in the middle track shows the borehole radius (RB) and invasion radius (RI); and the right track shows field resistivity logs (SRES, MSFL, LLS, and LLD). The right track also contains computed DLL logs under the given model Rt (RTSQ). As shown in the right track, the computed logs agree closely with the field LLD and LLS logs after inversion is completed. 10 m = 32.8 ft; 1 in. = 2.5 cm.

Contents

GeoRef

References

Related

Citing Books via

Close Modal
This Feature Is Available To Subscribers Only

Sign In or Create an Account

Close Modal
Close Modal