GAS READING WHILE DRILLING

2nd GCC-EU Advanced Oil and Gas Technology Conference Abu Dhabi.

ABSTRACT

HYDROCARBON GAS INTERPRETATION USING AN ADVANCED GAS DATA

ACQUISITION SYSTEM

Suresh Gadkari and Herve Chauvin, Geoservices S. A.

A basic and immediate requirement during drilling is accurate indication of formation fluid

type and saturation. For quite some time now, hydrocarbon gas recovered from drilling fluid

returns (ditch gas) has been used as an oil and gas indicator. Gas data interpretation, however,

has never assumed the status of an independent system for recognizing oil and gas zones.

Until recently, the equipment used for gas extraction and detection was not sufficiently stable

or efficient to provide reliable output. With the development of a constant volume degasser, a

much closer representation of the gas in the mud can be determined from the ditch gas values.

In addition, the introduction of improved detection systems has resulted in high resolution,

high speed, consistent analysis. These improvements enable gas data output that can provide

diagnostic properties. This is especially true for heavier components of the gases, which

make up a small proportion of the total, but are valuable as indicators.

There are many hurdles to overcome in gas data interpretation. Mechanical drilling

conditions, type of mud, mud additives, differential pressure, etc. cause variations in the

recorded gas data. Petrophysical properties, such as porosity, saturation, etc. are also

unknown at the time of drilling. Various methods have been developed to normalize gas data

but these are not sufficient for all conditions. It is therefore necessary to understand correctly,

and take into account, factors that influence the recovery of gas from the mud stream, as well

as the limitations of gas data interpretation.

The use of gas ratio analysis is one of the many tools that have been used effectively for real-

time gas evaluation. These ratios generally compare the relative quantities of the heavier

components with the lighter fractions, with different ratios corresponding to different

reservoir and fluid types. Analysis of the different combinations of gas fractions can lead to

fluid type identification and yield other significant information. Ratios bring out these

indications by enhancing the aspects that are not easily picked up by visual examination of

raw data. If such indications are available in real time, operators can reduce rig time and

expenses on wire-line logging, sampling, etc.

The ratios suggested here make this possible to a large extent. These ratios have been

validated with exceptional results in many basins of South East Asia. Quality data,

experienced personnel and careful application of scales are necessary for the effective use of

these tools.

A key first step for proper assessment is the definition of a clear format for data presentation.

Basic gas data, the ratios, and the variables that affect the data are all presented side by side.

This helps to bring out the salient features of the gas ratio curves. Final judgment regarding

fluid characterization and other aspects can be reached through the use of cut-offs and

comparisons.

2

HYDROCARBON GAS INTERPRETATION USING ADVANCED GAS DATA

ACQUISITION SYSTEM

Suresh Gadkari and Herve Chauvin, Geoservices S. A.

Introduction

Hydrocarbon gas released from crushed cylinder of rock is the first indication of the presence

of oil and gas in the zones being drilled. These indications are not just inferences, but direct

evidence of the presence of oil and gas. It is therefore imperative that this prime, cheaply

available parameter be used to maximum benefit.

Various gas detection methods have been in use for many years. Initially, only total gas was

recorded. Operators continue to use total gas, primarily for the purpose of hydrocarbon

potential and also for safety of operations. Later, chromatographic analysis became available,

and operators began to use the relative values of the various components to help determine

reservoir qualities. Unreliability of the gas data, however, meant that this direct evidence

could not be accepted independently as a trustworthy parameter. Deviations in the gas data

resulted from various factors, such as inconsistency of data acquisition and hole conditions at

the time of drilling.

If gas data is to be of any interpretative value, a basic requirement is that gas recorded should

be the same as the actual gas in mud. This aspect is of prime importance; all the efforts to

process the data by using ratios diagrams, charts or calculations are not indicative of changes

in formation unless the extracted gas itself is representative of true gas in mud. Recent

technological developments in this direction are significant.

Sensitivity, accuracy and consistency are necessary to obtain dependable interpretation

curves. Earlier degassers available to the market were not able to separate a sufficient

quantity of representative gas in mud. The conventional degasser may not be able to pick up

representative gas volumes, especially the heavier components, with consistency. This

inconsistency may distort the processed analytical curves, leading to difficulties in the

interpretation of gas data. With the newest generation of degassers, however, the gas released

from degassed mud is almost same as, if not equal to, gas in mud. Figures 1 and 2 show

effects of fluctuations in mud level on conventional and constant volume degasser.

Conventional degasser shows false decreases and increases in gas volume but constant

volume degasser is not affected by mud level fluctuations.

3

Trap Flooding :Upper part of the figure shows the effect of mud level fluctuations on the

out put from conventional degasser resulting in a spurious peak. Lower part of the figure

shows out put from constant volume degasser showing consistency with the fluid content

of the reservoir.

.

Fig.1

TG C1 C2 C3 iC4 nC4 iC5 nC5

TG C1 C2 C3 iC4 nC4 iC5 nC5

4

Additional developments that are changing the nature of basic gas data are improvements in

the gas analysis system or panel. The older generation of ―hot wire‖ and thermal conductivity

detectors have largely been replaced by the Flame Ionization Detector (FID).

Today, most reservoirs are drilled at a relatively high rate of penetration (ROP). In addition,

with the introduction of PDC bits there has been a great increase in average drilling rate

throughout the well. Thus a faster cycle time for chromatographic analysis is of utmost

importance. Decreases in retention time or chromatographic cycle time have resulted in better

resolution of the gas variations, and reduce the step-like appearance of the gas curves during

Trap starvation : upper part of the figure shows Gas data obtained from conventional

degasser affected by inadequate supply of mud to the degasser. Lower part of the figure

shows out put from constant volume degasser that is consistent with reservoir contents.

Fig.2

5

fast drilling. New-generation FID chromatographs also show better sensitivity and more

accuracy in the results. Both these aspects provide proper representation of heavier

components, which are very useful for interpretation of gas data. Enriched with all these

improvements we are now in a position to attempt interpretation of gas data. Upper part of the

Figure 3, out put from conventional degasser shows the effect of trap flooding on

interpretation curve, lower part of the figure is the interpretation curves from modern

equipment, eliminates spurious peak and gives much reliable curves because of better

resolution.

It is possible to make out oil zone and contacts from HM, LH and LM with the help of

better equipment. Upper part of the figure shows data from conventional equipment.

Lower part of the figure shows the output from constant volume degasser. Although basic

deflections are similar the curve from modern equipment is more reliable.

Fig.3

6

Gas Ratios

A direct plot of total gas and chromatographic curves is often used as an indicator of

hydrocarbons. It is not often easy, however, to differentiate oil and gas zones. The raw data

plot may not reveal its secrets, but when presented in the form of gas ratios it can be much

easier to detect reservoir character. The magnitude of significant variations in the gas plot is

small, hence it is necessary to subject these values to some formula which will enhance the

changes in the ratio curves, making them clear and easy to pick up.

There are many methods used for gas interpretation. Formerly, the only methods available

were gas composition diagrams from a single depth, such as triangular diagrams, Pixler plots,

etc. Later, the ability to produce a continuous plot of Wetness, Balance and Character (Wh,

Bh, Ch) ratios increased the practicality of gas ratios in reservoir interpretation, enabling

comparison of the gas ratios with the masterlog and electric logs. Additional ratio plots, such

as C1/Cn (Cn = C2, or C3, or C4, or C5 ) also were developed to aid interpretation.

Oil

Water

Gas

Gas

Gas

Ratios TG & CHR

Shallow Zones – Dry Gas

mainly C1

Mod, Dry Gas Zone

C1,C2,C3 Traces of C4 & C5

Wet Gas

Higher percentage of heavies

LM Light / Medium

LH Light / Heavy

HM Heavy / Medium

Water

Figure 4

Oil Zone

Deflection of three gas ratio curves in response to reservoir character under ideal conditions.

7

Gas Ratios - LM/LH/HM (Reservoir Fluid Determination)

The fluid type and saturation are the two fundamental aspects that require immediate

assessment while drilling. It is possible to display these ratios in real time. The indications

available in real time can help operators plan the wire line runs and sampling programs.

Recently, ratio plots consisting of three specific curves, viewed with the total gas curve, have

successfully been used to reveal fluid composition.

The ratios are:

LH – Ratio of Light to Heavy.

100 X (C1+C2) / ((C4 + C5)^3)

LM – Ratio of Light to Medium.

10 X (C1) / ((C2+C3 )^2)

HM – Ratio of Heavy to Medium.

((C4 + C5)^2) / C3

The curves LH and LM have lighter gases in the numerator; hence with an increase in density

of the hydrocarbons recovered, the curves deflect to the left (LH and LM decrease).

The curve HM places the heavy components in the numerator; thus with an increase in

hydrocarbon density the curve deflects towards the right (HM increases).

The basis for these equations is that the composition of the liberated gas varies with the type

of hydrocarbon content of the reservoir. A dry gas composition shows a very low percentage

of heavier gases such as C4 or C5, if these components are present at all. An increase in the

density of the hydrocarbons leads to an increase in the proportion of the heavier fractions, and

gases associated with other hydrocarbon fluids will contain a larger proportion of heavier

components. The density of the hydrocarbons in the reservoir will be reflected in the gas

composition recovered at the surface; thus the proportion of heavier gases increases from dry

gas to heavy crude oil.

With this set of curves, there are no limits suggested to link the deflections of the curves with

the type of hydrocarbon. Instead it is suggested that each section of the well be viewed

separately. A comparison of ratio deflections, combined with the amount of total gas for

hydrocarbon-bearing beds near each other, allows interpretation of reservoir content. The

limits on the deflection of the curves change with the type and properties of mud, and

according to petrophysical properties such as porosity, water saturation etc. It is therefore

necessary to judge the different sections individually.

In most cases where reservoir hydrocarbons are present, these two sets of curves cross over or

approach each other. The relative extent of crossover or approach can indicate the type of

hydrocarbon.

The scales for each ratio vary depending on ratio type. The variation is basically dependent

on type of mud and mud weight (MW), plus fluid type.

8

The following scale generally satisfies balancing MW conditions in synthetic oil-based mud

(SOBM):

LH LH and TG – Log scale 0.01 – 10000

HM gives better results on a linear scale 0 – 200

superimposed over log scale.

The scales suggested usually provide good results. Deviation from the expected range of

response is possible in some cases. A large increase in HM values and decrease in LM and

LH values suggests a large proportion of heavy components in the recorded gas; this condition

has been observed at some locations. In such cases it would obviously be necessary to

increase the HM scale, or decrease the LH and LM scales to fit the curves. Deflections of

these curves depend not only on hole and reservoir conditions but also on fluid character,

temperature etc. Hence it would not be proper to link directly the amount of deflection (by

giving absolute values) to the fluid type. Individual phases of a well, or rather each section

with identical drilling conditions, should be considered separately. In the cases of some fields

where most of the peripheral conditions are identical, it is possible to obtain a reasonable

estimate of both fluid type and saturation.

Figure 4 shows variations in response of these curves in different types of gases encountered

in zones of different composition under ideal conditions. The shallow dry gas zone will not

show any ratios, as medium and heavier fractions are absent. Moderately dry gas zones show

a decrease in the LM ratio; the heavier fractions are absent or are present only in small

quantities. Gas with heavier fractions will show up in the HM and LH ratios, depending upon

the amount of heavy components present. Oil or condensate zones show large deflections.

There is a difference in the response to the oil and gas zone. HM and LH show strong deflection

in case of oil while Total gas is more for gas zone. Water zone can be made out from very low

deflections.

Fig. 5

9

Lower part of Figure 3, out put from modern equipment shows the deflections in an oil zone.

Oil water contact can easily be made out from decrease in HM and increase in LH and LM.

It is necessary to take help of total gas to draw the inference as shown in the Fig. 5.

Deflections of all the three curves LH, LM, and HM are stronger in the oil zone. Lower gas

zone shows less magnitude of deflections but larger amount of total gas. Plot of total gas or

HCI on linear scale is also useful for determination of water cut-off.

The proportion of oil and gas in the zone is variable and will affect the response of the three

curves. Associated gas may show high HM and low LH and LM. This condition may be

interpreted as oil but a wire-line sample may show only gas, depending on the proportion of

fluids in the reservoir and their relative permeability.

Identification of source rock is possible with the help of LH, LM, HM curves. Figure 6 shows

shale section with increase in HM and corresponding decrease in LH and LM curves.

Decrease in LH is more than decrease in LM signifies increase in heavier components of gas.

Total Gas

Gas recorded in shale section with increase in HM and corresponding decrease in LH and LM

indicates source rock.

Fig. 6

10

HCI curve

This curve indicates oil saturation.

HCI – Hydrocarbon Indicator

HCI = ((2C2 + 3C3 + 4(iC4 + nC4)+ 5(iC5 + nC5))^0.5)/500

The HCI ratio gives better indications in water-base mud. It may be necessary to adjust the

scale for the curve or even formula depending on local conditions. Plot made on linear scale

gives better indications.

Variables in Gas Data Interpretation

There are many variables that create hurdles to be overcome in gas data interpretation. It is

necessary to take into account all these conditions, and their effects and corrections if

possible. Errors can occur if interpretation of gas data is attempted without taking into

account the factors that influence the basic gas data. Usually the quality of the gas data is

blamed for the inability to arrive at conclusive indications. Often, however, sufficient

understanding of the processes that have affected the basic data, or corrections for the specific

case, are not available. Following is a list of the variables that affect gas data.

Accountable (possibility of correction):

• Rate of Penetration

• Hole Size

• Flow Rate

• Degasser Efficiency

• Recycled fluids

Unaccountable (corrections not available):

• Differential Pressure (Mud weight )

• Mud type and viscosity

• Surface losses (loss of gas from bell nipple to degasser)

• Swabbing

• Surging

• Caving

• Diffusion

• Mud Loss/Mud gain

• Mud temperature

Petrophysical Properties:

• Saturation

• Permeability

• Porosity

• GOR

• Density of fluids

Corrections are available for some of these variables. Normalization of the gas values

corrects for hole size, rate of penetration, and flow rate. The influence of recycled fluids can

11

be assessed by measuring gas in and gas out. Degasser efficiency can be determined. Other

factors can influence gas data for which corrections are not available. Thus corrections can

not be applied to the effects of trap (degasser) starvation and trap flooding as shown in the

figure 1 and 2.

Differential pressure is the main parameter that affects the gas data. The amount of gas

recovered at the surface is only a fraction of the actual gas per unit volume in the reservoir

drilled. In addition, the proportion of gas components recovered is not the same as the actual

in-situ composition. The extracted proportion of the gas components depends largely on the

differential pressure. Higher differential pressure reduces mud gas content; in particular, the

heavier components will be reduced or absent.

The type of mud also affects the composition of the gas recorded. Water-base mud is

probably the best for gas recovery, whereas contamination of the mud with crude oil increases

retention of gas in mud and thus increases recycled gas. Recycled gas makes interpretation of

the gas data very difficult.

Effects of the loss of certain gas components are known but the estimation of the extent of

loss or gain is rather difficult. A phenomenon such as pre-flushing (a decrease in the amount

of gas in permeable oil- or gas-bearing zones) is not easy to predict from ROP and sample

description alone. It is necessary to take into account these factors before coming to any

conclusions.

Gas Ratio Analysis Log

A gas ratios log of consistent design has been developed for use in the field. Figure .7.

Fig. 7

The figure shows the suggested field format for Gas data Interpretation.

12

The log is divided into 8 sections; from left to right these are:

1. Drilling parameters, including

a) Mud weight

b) Hole size

c) Flow rate

d) ROP

2. Depth

3. Chromatographic Analysis

4. Lithology

5. Direct Florescence

6. Cut Fluorescence

7. Gas Ratios LH, LM, HM

8. Gas Ratios Wh, Bh, Ch

The parameters grouped in the first column affect gas output and as such these parameters

should be available for ready reference. A relatively heavy MW, for example, results in low

gas readings and affects the percentages of the individual components (C1—C5), and thus the

gas ratios. A large hole size at surface results in high gas values when compared to similar

formations drilled with a smaller hole size. An increase or decrease of mud flow rate will

affect gas output if all other parameters are constant.

Conclusion

In many observed cases, it is possible to differentiate formation fluids in real time by using the

LH LM HM ratios. Oil-water and gas-water contacts can be determined. Estimation of oil

saturation is also possible. Source rock can be identified. Difficulties can occur in judgment of

some zones as gas in mud is distorted due to well/drilling conditions. A comparatively high

percentage of oil in predominantly gas zones may also result in a misleading interpretation.

Further studies in both instrumentation and interpretation are necessary to improve

identification of reservoir properties.

The curves in the figures are based on the general observation of the authors of the study of the data acquired

using the ReservalTM constant volume gastrap and the ReservalTM high speed chromatograph analyzer both

exclusive Geoservices products. Fig. 7 is Geoservices Field format. The Ratios were formulated by Suresh

Gadkari of Geoservices.

Selected References:

1. Haworth, J.H., Sellens, M. and Whittaker, A. ―Interpretation of Hydrocarbon Shows Using Light (C1 – C5)

Hydrocarbon gases from Mudlog Data.‖ AAPG Bull., Vol.69, No. 8, 1985.

2. H.L. ten Haven, P. Arbin, B. Simon, G. Collo, J.P. le Cann and P. Mulero. ―Applications and Limitations of

Mud Logging Gas Data in the Detection of Formation fluids and Overpressure: Examples from South-East

Asia.‖ Gas Habitats of SE Asia & Australasia, Oct. 1998.

3. Salo, Jonathan Peter, and Eddy Luckiyanto, Conoco Indonesia Inc., Jakarta, Indonesia; and Suresh

Gadkari, Geoservices S.A., Paris, France. ―The Successful Characterization and Identification of In-Situ

Reservoir Fluid Content During Drilling Operations with SOBM (Synthetic-Oil Based Mud)‖, AAPG Poster,

Bali, Indonesia 2000.

4. D. Kandel SPE, Total Fina Elf: R. Quagliaroli, SPE, ENI Agip Div. G. Segalini, Total Fina Elf, B. Barraud,

Total Fina Elf, ―Improved integrated Reservoir interpretation using the Gas While Drilling (GWD) data.‖

SPE European Petroleum Conference Paris October 2000.