Go To Review

Evaluation of Geothermal Production for Sulfide Stress Cracking and Stress Corrosion Cracking

Dr. Russell D. Kane
CLI International, Inc.
Houston, Texas USA
email: rdk@clihouston.com

ABSTRACT

Operational problems were evaluated involving in-service failures of geothermal well production casing and wireline. Various aspects of the service condition and materials of construction were considered including: materials, grade, alloy composition, thermal stresses, temperature, chloride concentration and hydrogen sulfide concentration. The original casing materials were API 5CT N-80 and the wireline was made from carbon steel. Evaluation of the service conditions indicated that high thermally induced tensile stress could have combined with the presence of hydrogen sulfide in the produced fluids to result in sulfide stress cracking (SSC) of both the steel casing and wireline materials. Alternative casing grade steels were considered with higher resistance to SSC. High alloy wireline materials were also considered for improved performance with resistance to both localized corrosion and stress corrosion cracking. These material selections were made with the assistance of a computer based systems (SOCRATES^TM).

A series of geothermal wells were completed with additional wells planned. However, during early production, several problems were encountered. These included a relatively shallow casing leak in an interval at a depth of about 340 meters. Secondly, a series of wireline failures were experienced during routine temperature/pressure surveying. The wireline broke while pulling out of the hole at a depth of 1130 meters. The wireline broke at 800 meters leaving 330 meters of wireline in the hole. Prior to expanding operations or conducting further well logging an investigation was performed to examine the cause of these unexpected failures.

MATERIALS

The casing material was API 5CT N-80 grade material. This is a non-restricted yield strength grade of oil country tubular goods (OCTG) having a yield strength range of 80,000 to 110,000 psi (550 to 760 MPa). There are essentially no compositional limits except for sulfur and phosphorus and typically no hardness limitations associated with N-80 steels. N-80 tubulars can be manufactured by either normalizing or quenching and tempering. The wide variation commonly exhibited by this steel grade results in widely ranging properties such as strength, hardness, toughness and resistance to environmental cracking.

The wireline material used in well logging was a single strand standard 2.3 mm wire commonly referred to as "plow steel". This is a high strength patented wire based on a high carbon steel grade such as UNS G10650. The tensile strength of this material is typically in excess of 1200 MPa.

ENVIRONMENT

The initial test conditions for the geothermal well indicated a bottom hole temperature of about 350 C with wellhead temperatures ranging from ambient (30 C) to over 230 C after a little more than an hour of production. An analysis of the well gas cap composition was made and found to contain 1100 ppm H2S. At an estimated shut-in pressure of 17.3 MPa, this resulted in a hydrogen sulfide partial pressure of about 0.2 MPa (2.8 psia). This partial pressure of hydrogen sulfide is above the 0.0003 MPa (0.05 psia) limit indicated by NACE MR0175 for defining sour service conditions where sulfide stress cracking (SSC) could occur. SSC is a form of hydrogen embrittlement cracking commonly observed in high strength or high hardness steels in which the hydrogen is generated by the sulfide corrosion process on the metal surface.

The first consideration was to evaluate the well conditions to explain the occurrence of the casing leak. A downhole caliper survey showed only minor corrosion, pitting and erosion. An in-situ downhole examination performed with a fiber optic device indicated that the casing leak was the result of parted casing. The failure was a transversely oriented crack in the casing in a location adjacent to a threaded connection at the depth at which the leak had been previously determined.

THERMAL STRESS

Based on the well records and production history, it was theorized that thermal expansion and contraction of the casing in an unsupported interval may have induced excessive tensile stress. During heating, the casing string would tend to expand and upon cooling it would contract. Because the casing is typically cemented to the wall of the drilled hole, the casing becomes constrained. If the casing is not adequately constrained over a given interval, stresses can develop as a result of relatively rapid and high magnitude thermal changes associated with geothermal production (See Figure 1).¹

The stresses (S) from differential thermal expansion (or contraction) are indicated by the following equation:

where B is the thermal expansion coefficient (3.8 x 10^-6/^oC), dT is the differential temperature (267 C), and E is the elastic modulus for steel (2 x 10⁵ MPa). Based on this simplified analysis, it can be seen that upon heating the upper portion of the casing strength, compressive stresses (690 MPa) could be developed that were at least as high as or which possibly exceeded the yield strength of the material (550 MPa). This would likely produce compressive yielding. Then upon subsequent shut-in, the upper portion of the casing would cool to near ambient temperature again resulting in contraction of the tubing. The compressive yielding that took place during heat could result in excessive tensile stress upon cooling. This axial tensile stress would be additive with the normal mechanical stresses during shut-in when the internal pressure stresses are maximum.

The location of the failure also fits the abovementioned scenario since severe lost circulation was reported during drilling in an interval of over 100 meters at this depth. This indicates a region of over 100 meters around the failure which could have poor cement quality. This unsupported region would have been severely stressed as a result of the thermal cycles produced by the variations in production rate of the well. However, while this scenario would likely produce yielding or buckling of the casing it was not considered likely that it would produce a failure such as a parting of the casing without addition influences.

SULFIDE STRESS CRACKING POTENTIAL

Based on the scenario of high thermally-induced tensile stress in the steel casing, a second and more likely failure mechanism may have existed in the well. The combination of high axial tensile stress in the casing along with the high level of hydrogen sulfide may have produced conditions for SSC during shut-in of the well. Under these conditions, tensile stress in the upper portion of the well was at a maximum. This resulted from the thermal contraction of the casing and the internal pressure inside the casing. These effects would have been accentuated in the 100 meter interval if the casing was not adequately supported.

The shut-in conditions in the well (i.e. 30 C and hydrogen sulfide partial pressure of 0.2 MPa) constitute conditions where SSC could occur in susceptible materials. N-80 casing materials are not normally considered resistant to SSC since they typically do not have compositional, heat treatment, strength or hardness requirements designed to enhance resistance to SSC. This grade of steel would be particularly susceptible to SSC at the location of failure during shut-in of the well. Under this situation, very high tensile stress is likely as a result of the thermal contraction of the casing and its internal pressure. Susceptibility to SSC would be expected to be maximum in the highest strength/highest hardness material in the broad range N-80 steels are normally manufactured.

To reduce the possibility of future problems associated with SSC, the use of API L-80 casing is recommended. This material is similar to N-80 except that it is manufactured by a quenching and tempering process and has a more restricted range of yield strength of 550 to 655 MPa and a hardness maximum of HRC 23. These requirements exclude the most susceptible, high strength and hardness materials along with producing a quenched and tempered structure which results in the overall increased SSC resistance of the L-80 casing.

WIRELINE FAILURE

The problems identified for the N-80 steel casing indicated that SSC was the most likely cause of failure. This was the result of the high hydrogen sulfide partial pressure in the geothermal well environment. Similarly, the most likely explanation for the wireline failures was also SSC. The very high strength wireline steel would be expected to be even more susceptible to SSC than the N-80 casing. As a result of this observation, the wireline material was changed to coldworked UNS S031600 (a austenitic stainless steel). The immediate observation after making this switch was the elimination of shallow wireline failures.

The only concern was that UNS S031600 while very resistant to SSC can be susceptible to stress corrosion cracking (SCC) in the presence of high temperature chloride solutions. In most cases, SCC of austenitic stainless steels is expected to occur at temperatures in excess of 60 C. This experience is highly biased toward service in aerated, chloride-containing environments. In this case, however, the service conditions associated with geothermal production are basically anaerobic meaning that the environment is basically deaerated. Under these conditions, the severity of SCC is typically lessened since the amount of localized corrosion is significantly lessened which, in turn, greatly reduces the possibility of initiating SCC.

In the present case, no analysis of the produced fluids were available. However, for other geothermal operations it was found that geothermal brines more typically have concentrated chloride environments and also may have buffered solutions as a result of the presence of high bicarbonate levels. Based on these assumptions, a predictive model for the selection of corrosion resistant alloys (SOCRATES^TM) was used to obtain possible serviceability limitations for this material. Based on assumed downhole conditions of 0.2 MPa hydrogen sulfide, 100,000 ppm chloride and pH 5.5, a maximum service temperature for UNS S-31600 was estimated to be around 100 C. Therefore, to minimize problems associated with SCC of UNS S031600, the use of this material should be limited to conditions in the well of lower temperature (i.e. only after shut-in and well cool down and very short durations at temperatures above 100 C. Removal of the end of the wireline after periods of high temperature service are recommended to further reduce the possibility of SCC resulting in failure in a subsequent run.

Additionally, for wireline operations at higher temperature, the use of a super duplex stainless steel or high alloy austenitic material is also recommended. The computer model indicated that candidate materials would be those with a pitting resistance equivalent [PRE = Cr + 3.3Mo + 16N = 1.5(W+Nb)] of 30. This would limit the wireline materials to the following commonly available materials: Duplex stainless steels - SAF 2507 (UNS S32750); High alloy austenitic stainless steels - Sanicro 28 (UNS N08028) and 254MSO (UNS S31254). Nickel-base alloys such Alloy 825 (UNS N08825), G-50 (UNS N06950) and C-276 (UNS N10276) would be expected to give even higher resistance to SCC for environments that had higher hydrogen sulfide levels and/or lower bicarbonate levels and pH.

Based on the results of the investigation presented herein, the following conclusions were made:

1. The failure of the N-80 casing was caused by a combination of problems involving

1. high thermally induced tensile stresses.
2. high hydrogen sulfide partial pressure.
3. the use of N-80 casing.

1. The mechanism of casing failure had all of the characteristics of sulfide stress cracking.

1. An alternative L-80 material was recommended for future completions which has higher resistance to SSC.

1. The steel wireline failures were also likely to have been caused by SSC.

1. The use of UNS S031600 wireline minimized the failure due to SSC, but there still exists a likelihood of failure due to SCC due to the presence of high temperature and chloride levels in the geothermal production.

1. Existing and alternative wireline materials were evaluated using a computer model. Temperature limits for UNS S31600 was estimated to be around 100 C. For higher temperature service, duplex stainless steels and high alloy austenitic stainless steels were identified with higher resistance to SCC. If even more severe conditions were encountered, nickel base alloys would be required.

1. Snyder, R.E., "Geothermal Well Completions: A Critical Review of Downhole Problems and Specialized Technology Needs", Paper No. 8211, SPE Annual Conference, September, 1979.

(0)