SOURING OF NEW IRIAN JAYA WELLS TRACED TO INDIGENOUS BACTERIA

P.J.B. Scott, Michael Davies
Cariad Consultants
Heraklion, Greece

Analysis using stable isotopes of sulfur combined with more traditional analyses demonstrated that souring of oil produced from several new wells in Indonesia was caused by indigenous sulfate-reducing bacteria.

The new wells are in an oil and gas field being developed on Salawati Island off Irian Jaya (Fig. 1) by the joint operating body (JOB) Pertamina-Trend Salawati, under a production sharing contract between Pertamina and Trend Kepala Burung Ltd., a subsidiary of Santa Fe Energy Resources Inc.

This area of Irian Java already has a number of producing fields in the tropical coastal jungle (Fig. 2).

The new development began in 1991 with the discovery well and two appraisal wells that tested oil, gas, and water.

Production from the three wells (Matoa-1, Matoa-2, and Matoa-3) was sour, with hydrogen sulfide (H2S) in the range of 1,850-8,000 PPM.

A sophisticated new technique using stable isotopes of sulfur combined with more traditional detailed biological and chemical analyses determined that the souring was caused by indigenous sulfate-reducing bacteria (SRB) with low activity, acting slowly over time.

Downhole conditions in the wells are marginal for the development of SRB; therefore, traditional techniques alone are not adequate to detect sufficient activity to find the cause for souring.

Metallurgical and corrosion rate determinations were used to select appropriate downhole and above ground materials and predict the life of the existing equipment.

The field's development calls for drilling 25 producing wells. Reinjection of gas is being studied as a means of maintaining reservoir pressure to maximize oil recovery.

Hydrogen sulfide gas is, therefore, a significant problem, not only from the safety and corrosion standpoint, but later, for maximum efficient recovery of oil and gas. By knowing the cause of souring, the best method of handling the sour conditions can be determined.

There are several different ways H2S can be generated in a formation and cause souring of the well. The following are six geochemical1 or biological causes:

Reduction of bisulfite chemicals, used as oxygen scavengers

Thermo-chemical sulfate reduction, usually at temperatures greater than 100 C.

Thermal decomposition of organic sulfur compounds

Reductive dissolution of pyritic material FeS2

Nonoxidative dissolution of pyritic material

Sulfate reduction by bacteria (Fig. 3).

Bisulfites were not used in these wells, therefore No. 1 is not a potential cause. This leaves the possibility of either geochemical (Nos. 2-5) or biological (No. 6) as being responsible for the souring in these newly drilled wells.

As part of the study to evaluate the SRB and H2S environment of this field, sulfur isotope analysis was applied to distinguish between these possible sources of H2S.

The study program was designed to ensure that the maximum recovery of reserves is undertaken in the most cost effective and efficient way.

INITIAL INVESTIGATIONS

Initial investigations of the test wells included a review of operating conditions (Table 1) and bacterial culture of three produced water samples from the storage and production tanks of Matoa-1.

About 105-106 SRB were found. This indicated a possible biological origin for the souring.

To determine if the SRB infestation in well Matoa-1 was "near well bore" or "far well bore" a hypochlorite soak treatment had been carried out prior to this investigation. The 3.3% calcium hypochlorite treatment was made with filtered pond water and injected downhole.

The conclusions from the treatment were that the H2S infestation was "far well bore" and caused by the presence of SRB. It was not determined whether the drilling program introduced the SRB or activated dormant SRB existing in the formation. After reviewing the hypochlorite procedure and results, it was clear that an alternative interpretation was possible. This interpretation was that a chemical reaction was occurring between the calcium hypochlorite and H2S, in addition to any possible biocidal action on SRB. Part of the reason for doubting that the source of H2S derived from biological activity on the basis of this test alone was the relative rate at which the H2S and the bacteria returned to their presoak levels after the soak. The H2S apparently returned quickly while the bacteria levels lagged behind and only reached 100-1,000/ml. Bacterial numbers in the 100-1,000/ml range are not generally thought to be sufficient to produce significant quantities of H2S.

The chemical reactions proposed were:

H2S + 2Ca(OCl)2 = H2SO4 + 2CaCl + Cl2H2S + Cl2 = 2HCl + SH2SO4 + CaCl + Cl2 = CaSO4 + 2HCl

These reactions were experimentally simulated. An approximately 3% solution of calcium hypochlorite was prepared and partly analyzed. The results are designated "initial" in Table 2.

Gas containing high levels of H2S from the gas treatment plant was bubbled through the solution for about 1 min and the solution reanalyzed. These are shown as the "final" results in Table 2. The conclusion is that during the hypochlorite soak, free chloride and hypochlorite were reacting with H2S to form acids, sulfur, and salts. This consumption of H2S by chemical reaction would account for the reduction in H2S measured during the treatment. Bacteria, if present, would also be reduced in number by the biocidal action of the hypochlorite, liberated chlorine gas, and acidification, thus reducing H2S production. It is not possible to conclude from the treatment that these two effects on H2S reduction were dominant.

BIOLOGICAL ANALYSES

To determine whether appropriate conditions exist for the growth and activity of microorganisms, produced waters were tested for important biological nutrients and environmental parameters (Table 3).

Results clearly indicated that sufficient nutrients exist to support populations of SRB in all three wells.

The presence of SRB in all three Salawati wells was indicated by culturing in selected media the produced waters and tubing samples for a variety of corrosion-causing bacteria, including sulfate-reducing bacteria (Table 4). Their numbers and activity, however, are relatively low.

Most of the SRB cultured are mesophyllic, that is they can be cultured only at temperatures below those found in the wells.

There was negligible growth of thermophiles in the culture media, although they may be present and able to grow at the temperatures under pressure conditions found downhole. Nonetheless, lack of growth under culture conditions does make it unlikely that SRB are highly active. H2S generation would, therefore, be slow and may have occurred over many years.

These results imply that the SRB were indigenous to the formation and were not introduced during drilling, although more may have been added. Growth of SRB was generally greater in the API RP 38 medium than in the modified API (MAPI), indicating that halophilic species of SRB are not abundant.

Any quantification of planktonic (nonattached) SRB, should be interpreted with caution because planktonic populations are not always indicative of sessile (attached) biomass, and may under-estimate the number of SRB by many orders of magnitude.

Because planktonic counts are unreliable indicators of SRB activity, two other methods were also used. These are scanning electron microscope (SEM) and EDS examination of corrosion products and surfaces of well equipment and stable isotope determination.

MICROSCOPY

Corrosion products and production tubing samples were examined in the SEM. Samples were fixed in a 2% buffered glutaraldehyde immediately after collection. Then, the samples were post-fixed in osmium tetroxide, critical point dried, mounted, and gold coated before examination in the SEM.

No biofilms were found on any of the samples examined. Bacteria present were isolated and rare.

SULFUR ISOTOPE

Natural sulfur contains four stable isotopes: 32S, 33S, 34S, and 36S, of which 32S is the major component (about 95%) and 34S the next most abundant (about 4.2%). The different isotopes vary in mass and therefore react slightly differently in most biological reactions, as well as in some physical processes such as diffusion and evaporation. In chemical reactions, lighter isotopic bonds vibrate at higher frequencies and are more readily broken.2 During bacterial reduction of sulfate, lighter isotopes of sulfur are preferentially converted to sulfide, resulting in a product that is enriched in 32S compared to geochemical sulfur, while the residual sulfate is enriched in 32S. These differences are detected by mass spectrographic analysis and usually expressed on a d34S scale, according to the following equation:d34S in% = [(34S/32S sample)/(34S/32S standard) - 1] x 103 The 34S/32S is the ratio of the number of 34S to the number of 32S atoms in the sample or the standard.

The international standard is troilite (FeS) from the Canon Diablo meteorite. Isotopic discriminations during measurement occur in both the numerator and denominator and therefore cancel, permitting determinations with a reproducibility of better than 1/10,000.

The natural isotopic fractionation indicates the speciation of the sulfur in a given specimen. The process that results in the most significant isotopic fractionation is sulfate reduction by bacteria. SRB can also reduce other sulfur species, such as sulfite.

The degree of sulfur isotope fractionation is dependent on the sulfur species reduced and is inversely related to the reduction rate per unit cell (the degree of biological activity) that results in some variability in measurement of isotopic fractionation.

The previous principles were applied to test the hypothesis that H2S generation in the Salawati wells has been produced by biological sulfate reduction. If bacteria were responsible, the H2S sulfur should be lighter, i.e., depleted in 34S, relative to the SO4^2- sulfur. If the H2S were generated thermo-chemically, this fractionation would not be apparent. Pairs of samples were analyzed, hydrogen sulfide and sulfate from produced water from Matoa-1 and Matoa-3. The results show that there is a large fractionation, with H2S sulfur considerably depleted in 34S (a smaller number) relative to the SO4^2- sulfur (Table 5). Samples from sulfur source "HS-" in Table 5 give the fractionation of sulfur from H2S. Where "SO4^2-" is indicated, the sulfur fractionation from aqueous sulfate was measured. Light isotopes are much more abundant in the H2S than in the residual sulfate of the produced water. The results indicate clearly that SO4^2- reduction by bacteria has occurred, rather than high temperature thermo-chemical generation. In all cases, the sulfur from the gas is lighter than from the residual sulfate. Furthermore, the d34S for Miocene marine sediments (the formation rock) is normally about +21. Because this is the original source of sulfur, this value is intermediate between the product of bacterial reduction (the gas fraction) and the residual aqueous sulfate. Although these results point clearly to a bacterial source, no information can be gained by this method on the rate or time of H2S generation. It is possible that the H2S was generated in the geological past (millions of years ago). This is unlikely, however, because the H2S would probably have reacted with metallic ions to form stable sulfides, such as iron sulfides.

METALLURGICAL CHECK

Samples were taken from various depths of the N-80 production tubing from Matoa-1. Longitudinal sections were taken through each of the threaded areas (inside and outside surfaces examined) together with cross sections through unthreaded areas. Samples of production piping were also taken and examined (Fig 4).

All of these samples were prepared and examined using standard metallographic techniques. The metallurgical structure consists of fine carbides in a ferrite matrix and is typical of a normalized plain carbon steel.

Some surface corrosion was detected on all the threaded surfaces (the outside of the tubing), with more corrosion on the inside of the tubing.

The extent of this general attack is least at the bottom of the string. This could be due to:

The temperature effect whereby general CO2 corrosion decreases above about 60 C.
Difference in the composition of the tubes
Differences in amount of alkalinity present.

Whatever the cause, the apparent differences in corrosive attack are slight and the extent of attack on all tubes examined was minor.

Selected areas of scale on the tubing were examined using EDS. These analyses indicated that the scale on the inside of the tube is largely sulfide. In some areas, chlorides have been playing a role in the corrosion reaction.

The outside scales were original millscale, only partially modified by exposure to the environment.

FIELD DEVELOPMENT

The analysis of stable isotopes has been used for a number of purposes, including tracking sources of pollution in the environment. This new application was able, for the first time, to identify bacteria definitively as the source of souring in newly drilled test wells in Indonesia.

A combination of this modern technique, coupled with more traditional techniques of bacterial culture, condition monitoring, and nutrient analysis has indicated that SRB are still moderately active in the formation and must be considered in development of the field.

The corrosion measured to date is not excessive, and existing well completion materials are generally acceptable.

A program has been instituted to monitor conditions in these wells to ensure that conditions do not deteriorate and cause renewed corrosion or excessive production of hydrogen sulfide. Precautions are being taken to avoid injection downhole of bacteria or nutrients.

At this stage, downhole biocide application is not recommended but this will be considered as part of the review of the results from the monitoring.

ACKNOWLEDGMENTS

The authors would like to thank JOB Pertamina-Trend Salawati for assistance in undertaking this investigation and for permission to publish this article. The assistance of Roy Krouse in performing the sulfur isotope analysis is also gratefully acknowledged.

REFERENCES

Marsland, S.D., Dave, R.A., and Kelsall, G.H., "An electrochemical approach to oil reservoir souring," Trans. IChemE. 68A, 1990, pp. 357-64.
Krouse, H.R., "Environmental sulphur isotope studies in Alberta: A review," prepared for The acid deposition research program by the Department of Physics, the University of Calgary, 1987, 89 pp.