Improved High-Pressure, High-Temperature (HPHT) Materials Qualification Using Dissolved H₂S Concentration as the Sour Service Scalable Metric

Relevant sour service standards still associate H₂S partial pressure (P_H2S) with SSC^1-4  and this parameter continues to be used to qualify materials for HPHT wells. However, there is recent evidence that SSC susceptibility correlates better with H₂S fugacity (f_H2S) and/or C_H2S at elevated total pressures.^5-7  The recent publication of ISO 15156-1:2020 now allows users the option to define environmental sour severity either by f_H2S, activity (a_H2S), or C_H2S in addition to P_H2S. These alternative sour severity metrics incorporate the thermodynamic nonideality effects on H₂S partitioning between gas/liquid hydrocarbon and aqueous phases at elevated pressure, temperature, and composition. The challenge now shifts to identifying a quantitative H₂S parameter that is scalable with total pressure from ambient to HPHT conditions. The current best two candidates for HPHT applications are either f_H2S or C_H2S. As both f_H2S and a_H2S are derived from the chemical potential (μ_H2S) and fugacity can be defined in each phase, it is sufficient to consider P_H2S, C_H2S, and f_H2S only.⁸  Figure 1 shows two Venn diagrams describing (a) all the factors that influence SSC and (b) the individual environmental factors that influence each sour severity metric.

While it is widely accepted that defining sour service limits based on P_H2S is problematic, there is no universal acceptance of an alternative metric. There has been significant progress in understanding the thermodynamics of H₂S under HPHT conditions,^5,9-12  but there are only a handful of studies linking this nonideal thermodynamic behavior to actual SSC outcomes.^9-10  Last, proper risk-benefit analyses of adopting an alternative modern H₂S-SSC correlation metric are scarce.¹³ 

The critical issue is therefore to determine which parameter, f_H2S or C_H2S, best characterizes higher-pressure sour systems for material qualification purposes. On the one hand, several authors^5-7  discuss differences in H₂S solubility in high-pressure vs. low-pressure systems and the relationship between concentration, fugacity, and activity. No evidence of the controlling SSC mechanism was provided. On the other hand, Krishnamurthy, et al.,⁷  recognized the impact of C_H2S-pH on the “mechanistic” control of SSC with respect to existing evidence and empirical modeling, but it is not a mechanistic assessment that proves the role of C_H2S in SSC. Regardless, once the industry arrives at a consensus on the preferred sour severity metric, and a precise quantification of the parameter is demonstrated, its significance to SSC testing outcomes with similar accuracy will need to be established.

To further the discussion, this paper addresses the merits and challenges of the two modern approaches to defining sour severity: f_H2S and C_H2S. On the one hand, the computationally derived f_H2S contributes to the equilibrium distribution of H₂S between the nonaqueous and aqueous phases, which influences sour severity at the metal/solution interface. On the other hand, the experimentally verifiable C_H2S has a direct correlation to hydrogen ingress into the metal that is a precursor to cracking. We consider C_H2S as a practical, experimentally verifiable approach for material qualification in mildly sour HPHT environments, and could be used to validate alternative computationally based/nonexperimentally verifiable approaches (e.g., f_H2S and a_H2S). The following literature review is intended to provide clarity on this important subject and provide a logical path forward to improving material qualification testing for HPHT service conditions.

Performing actual NACE-compliant H₂S fitness for service qualification tests at full wellbore pressure is not practical for most projects. Sour service autoclaves that can safely maintain a pressure differential of >10,000 psig for 720 h are rare, costly, and have limited internal volumes. Therefore, to reduce costs and to avoid the complexities, the Standard permits sour service testing in low-pressure laboratory equipment provided that the candidate material is qualified in a suitable environment that is at least as severe as the field environment.¹  This is generally accomplished by using a “test gas-containing H₂S and CO₂ at the same partial pressures at the intended service.”¹  This approach ignores the influence of high pressure on the cracking behavior of susceptible materials. The consequence of this policy, while not fully understood at the time of adoption in 1975, results in a conservative, or possibly under-conservative, design philosophy.¹³  The intrinsic over-conservatism of the above NACE/ISO policy becomes consequential when selecting materials for mildly sour HPHT wells with wellbore pressures >10,000 psig.^5-6,14  (The policies leading to over/under-conservative design will be addressed in the following paragraphs.)

One challenge is defining the appropriate safety margin in an SSC experiment. No standard exists to define minimum safety factors for SSC testing. Table 1 provides a proposed quantitative definition of “degrees of conservatism,” although by no means are these values absolute. It is hoped that the recent investigations examining the modern sour severity metrics will lead to a consensus.

The issue is confounded further because materials may be qualified for sour service at 14.7 psia (101.3 kPa) with “H₂S at the partial pressure of [the] intended service.”¹  Performing laboratory tests at atmospheric pressure, while maintaining the field P_H2S, requires increasing y_H2S. For example, a 20,000 psig HPHT well containing y_H2S = 2.5 mol ppm yields a pseudo-P_H2S of 0.05 psia.¹⁵  At atmospheric pressure, to maintain the 0.05 psia P_H2S the test gas must contain y_H2S = 3,400 mol ppm (see Table 2). If the solubility of H₂S remained constant with total pressure (i.e., the gas solubility is linear with pressure) then substituting total pressure with a higher y_H2S in the test gas would be inconsequential. This not the case. H₂S solubility decreases with increasing total pressure by at least an order of magnitude (refer to Figure 14 as an example). Thus, qualifying materials based on equivalent P_H2S may lead to over-conservatism particularly when qualifying materials for service conditions with total pressures >10,000 psig (69 MPa).^5-7,12 

In addition to over-conservative design concerns, the Standard also allows for a simplification for systems whose total pressure exceeds the bubble point/saturation pressure (e.g., “gas-free wells”). The assumption is made that the severity of the environment at pressures above the bubble point can be characterized by the P_H2S calculated at the bubble point, which may be nonconservative when considering the modern f_H2S and C_H2S approaches.⁸  However, with modern ionic-equation of state (EOS) frameworks readily accessible to all users, f_H2S, a_H2S, and C_H2S can be calculated at any temperature and total pressure, rendering the bubble point approximation obsolete.

Clearly, P_H2S is not an appropriate metric for HPHT material selection and design, because it over predicts the true sour severity (be it f_H2S or C_H2S) of the actual field conditions. Defining material sour service limits strictly based on P_H2S implies that the effect of total pressure on cracking susceptibility is inconsequential. The premise that material qualification laboratory tests, conducted at atmospheric pressure, can be designed merely by decreasing the total pressure and increasing the y_H2S to maintain the same P_H2S expected in the field and yield an identical SSC outcome for the same oilfield material is flawed. High-pressure laboratory tests and field experience confirm that the risk of cracking is significantly lower than would be expected based on laboratory tests results performed at atmospheric pressure with the same P_H2S.¹⁶  Therefore, resolution on the feasibility of implementing either f_H2S or C_H2S is an important contribution to improving material qualification testing for HPHT wells, although it is not the sole consideration for users.

Nevertheless, the industry needs to adopt a new realistic design metric that is fully scalable with total pressure between atmospheric pressure and 20,000 psig. An improved sour severity metric to describe HPHT environments does not imply that other environmental (including mass transport, pitting, or film formation) and/or mechanical factors are not important contributors to SSC. Rather, the authors argue that P_H2S should be replaced with C_H2S as the new sour service metric.⁽²⁾ Preference for C_H2S does not preclude the option to select f_H2S or a_H2S as an alternative sour severity metric, pursuant that the thermodynamic calculations are performed correctly, if so desired. The following sections provide the history and rational for adopting C_H2S as the sour severity metric.

In the classic SSC scenario, the source of H₂S originates from the oil or gas phase(s) and then partitions from the nonaqueous phase into the aqueous phase. Within the aqueous phase, the dissolved H₂S (C_H2S) then migrates to the metal surface based on mass transport properties (i.e., velocity effects) preventing recombination of atomic hydrogen from proton reduction to molecular hydrogen, thereby forcing atomic hydrogen uptake into the metal. From a cracking mechanism perspective, it is either C_H2S^5,17  or the f_H2S^9-10,18  that influences SSC at the crack tip, rather than the P_H2S. For convenience, however, the Standard categorizes SSC susceptibility based on P_H2S as a surrogate for SSC severity induced by H₂S.

The Standard partial pressure thresholds implicitly presume that the P_H2S is proportional to C_H2S via the Strict Henry’s Law, which assumes ideal gas behavior:

where the ideal (or strict) Henry’s Law constant for H₂S is ^Ik_H2S = 230 mg/L/psia (0.1 mol/L/atm) for freshwater at 75°F (24°C) and 14.7 psia (1 atm).^19-21  (Refer to “Nomenclature” for all other symbols.) P_H2S is a relatively accurate approximation for C_H2S at room temperature and low total pressures for dilute solutions. Irrespective of the total pressure of the system, however, the sour service limits for all materials specified in the Standard are defined in terms of the partial pressures of H₂S and CO₂. The Standard assumes that the ideal gas behavior governs and that the influence of the hydrocarbon phase on the cracking susceptibility is negligible and can be ignored.

The historical practice of extrapolating the strict Henry’s Law to the full wellbore total pressure was suitable and afforded reasonable safety factors for low-total pressure systems. As HPHT wells have become more prevalent and the total pressures continue to rise, the safety margins have increased significantly to the point that they are onerous and impractical based on simple P_H2S calculations. While a sour service limit of 0.05 psia H₂S may be appropriate for wells with low total pressures, the authors are not aware of any evidence demonstrating that 0.05 psia (0.3 kPa) H₂S ought to be the sour service limit between 10,000 psig (69 MPa) and 20,000 psig (138 MPa) at 75°F (24°C).

Recognizing the limited relevancy of the historical “regions of environmental severity” to modern HPHT applications, various agencies involved in the maintenance of the Standard have agreed to the recent publication of ISO 15156-1:2020 which permits users the option of implementing P_H2S, f_H2S, a_H2S, or C_H2S as metrics to define environmental severity. For clarity, Table 3 categorizes the modern sour service metrics based on phase behavior and whether the parameter is directly measurable in the lab or is derived from rigorous computational modeling. While the adoption of ISO 15156-1:2020 is an important contribution and milestone, the standard does not provide clear guidance on the validity of, or how to validate, the alternative sour severity metrics.

P_H2S still remains the de facto metric to define sour severity for a variety of reasons including historical experience, safety concerns, limited available data, and/or logistical issues with capturing and preserving field samples. To safely reduce over-conservatism for material qualification in mildly sour environments at HPHT conditions, the industry ought to adopt a universally accepted P_H2S alternative that is empirically based and is fully scalable across the entire service domain. Of the metrics recognized by the Standard, a concentration-based metric is the only metric that is directly experimentally verifiable. While both f_H2S and C_H2S can be used to translate historical SSC material testing experience (generally obtained at total pressures <5,000 psig [34.6 MPa]) to predict SSC behavior between 10,000 psig and 20,000 psig (69 MPa to 138 MPa), C_H2S allows the user to explicitly set their own margin of safety. Please refer to Table 4 for a summary of additional desired properties in a modern sour severity metric.

Having demonstrated the limitations of using P_H2S to predict H₂S behavior at elevated pressures, the following sections provide a review of industry’s attempt at implementing alternative modern approaches to better correlate the nonideal behavior of H₂S within the HPHT domain to actual experimental data.

The observed dissolved H₂S “saturation plateau” arises from the fact that H₂S does not partition equally between the hydrocarbon and aqueous phases because H₂S has limited solubility in water. Literature suggests that the “saturation plateau” of C_H2S is related to the effect of increasing pressure on the liquid and the gas. Therefore, for a given y_H2S, C_H2S increases with increasing total pressure until an apparent “saturation plateau” is reached. Within this apparent “saturation plateau” region, the C_H2S is nearly independent of total pressure;¹⁰  furthermore, it is presumed that this observation is a universal phenomenon.

Nelson and Reddy applied the Ensemble Henry’s Law approach to compare the phase behavior of a well at 10,000 psi (69 MPa) total pressure containing 1 mol% H₂S (100 psia [690 kPa] H₂S) and a low-pressure laboratory fitness-for-purpose (FFP) test containing exclusively 100 psia H₂S.¹²  For their scenario, the FFP test contained 11 times more dissolved H₂S than predicted in the actual well. The authors concluded that the FFP test conditions were significantly more severe than the environmental conditions anticipated in the well.¹² 

Based on the iso-fugacity criterion, which is derived from the concept of equal chemical potentials (i.e., ⁠), Grimes, et al., have argued that “the equivalent fugacity between a field and laboratory condition can be matched irrespective of the nature—gas or liquid—of the phase present in the field.”¹⁰ 

Thermodynamically, the fugacity of a gas is its effective, or chemically equivalent, analog to the physical partial pressure.³²  Fugacity incorporates the nonideal thermodynamic effects in the real gas that are ignored in the partial pressure derived from the ideal gas law.⁸  The f_H2S can be expressed in the units of pressure based on the following expression:

The ϕ_H2S value is a pressure, temperature, and compositionally dependent parameter. It is a calculated, thermodynamic coefficient that cannot be directly measured in the lab, but rather the EOS model parameters are fitted to experimental data.³²  Computational modeling is currently the most commonly used approach to determine fugacity in practice as physical testing is quite expensive.

Should the industry adopt f_H2S as an improved sour severity metric over P_H2S, the user should consider the following discussion points:

1. Under liquid-full conditions, there is evidence that f_H2S is a better correlation to SSC outcomes than P_H2S, and may alone be sufficient to describe the system.^10,18 

2. Neither f_H2S nor ϕ_H2S can be measured directly; they must be calculated. This may lead to inconsistencies when comparing ϕ_H2S calculated from different EOS models, which are often calibrated against experimental data. Currently, there only a handful of publicly accessible H₂S-H₂O-salt datasets available.^{24-25,29,33-38} 

3. Figure 8(a) indicates that as the temperature increases, ϕ_H2S approaches unity (ideal behavior) at elevated total pressure. Consequently, because there is only a marginal reduction in conservatism based on f_H2S, P_H2S alone may be sufficient to assess sour severity at elevated temperature and total pressure.

4. Fugacity does not directly predict H₂S solubility in water: (a) the effects of high pressure (Figure 6)^26,30  and (b) the effects of salinity (see the Ensemble Henry’s Law Approach [H₂S Partitioning Between the Nonaqueous and Aqueous Phases] section and Figure 10 for discussion).²⁸ 

5. A single fugacity value can correspond to multiple C_H2S for hydrocarbon mixtures containing different y_H2S resulting in potentially different SSC outcomes. (To be discussed in further detail in the Application of The Modern Sour Severity Metrics section.)

6. To maintain the same level of conservatism afforded by f_H2S, the user would have to replicate the field conditions exactly in the lab at full wellbore pressure. This is complicated because evaluating materials in mixed gas/oil compositions and at total pressures between 10,000 psig and 20,000 psig is difficult, may unnecessarily increase testing costs, and is generally not practical for most routine laboratory applications. For example, most labs use nitrogen as a nonflammable substitute for methane. While holding y_H2S constant, substituting nitrogen for methane as the balance gas results in higher ϕ_H2S (see Figure 8[d]). The solubility of H₂S in the “gas” has a significant effect on the solubility of dissolved H₂S in the test brine.

In summary, we argue that a fugacity-based approach is applicable for many sour oilfield applications because the actual C_H2S is much lower than predicted by P_H2S/f_H2S as demonstrated by Nelson and Reddy¹²  and Kumar, et al.²⁹  We agree that f_H2S provides a better representation of field conditions in the lab, which may induce more or less conservatism than classical P_H2S. In addition, fugacity is particularly useful when assessing low-temperature and high-pressure environments, where ϕ_H2S is predicted to be <<1. However, fugacity is not the only valid design approach when assessing sour severity under HPHT well environments.

The advantage of the Ensemble Henry’s Law is that currently available ionic-EOS frameworks^{9,29-31,39,41-42}  incorporate the three correction terms to predict H₂S behavior under HPHT conditions. (Incorporation of three correction terms may make the thermodynamic calculations more complicated to compute, and thus subject to concomitant inaccuracies. However, unlike f_H2S, the predicted C_H2S can be independently verified in the lab to validate thermodynamic predictions.) When using these models, it is important to consider that, based on the derivation of the Ensemble Henry’s Law,²⁶  the fugacity and activity contributions are inseparable when predicting gas solubility.

The Ensemble Henry’s Law approach is a significant improvement over the Strict Henry’s Law and fugacity-corrected approaches. While ionic-EOS frameworks are gaining wide industry acceptance, the models have not been fully validated within the sour HPHT domain of interest. Relying on ionic-EOS frameworks without experimental verification at total pressures >10,000 psig is a concern for the following reasons:

1. Current ionic-EOS frameworks combine two different thermodynamic theories to describe HPHT systems: The hydrocarbon phase is described by a fugacity-based algorithm (derived from van der Waals equation), while the aqueous phase behavior is described by an activity-based algorithm^26,30  (an extension of the Debye-Huckel theorem) coupled to a Poynting correction factor.³²  Both algorithms use different thermodynamic parameters and incorporate numerous complex equations.⁴³  Calculations converge when the chemical potential in the nonaqueous phase (⁠⁠) equals the chemical potential in the aqueous phase (⁠⁠). The useful thermodynamic parameters (i.e., γ_H2S, ϕ_H2S, and exp[ξ]) are derived from the chemical potential, a derivative of Gibbs free energy.²⁶ 

2. EOS models often regress thermodynamic parameters from simple, multi-dimensional systems. Unlike the CO₂-H₂O-salt system,⁴³  there are only a handful of reliable H₂S-H₂O-salt databases available,^{24-25,29,33-38}  and the majority of the public datasets contain experimental data up to 5,000 psig.³⁰  The authors are not aware of any published, experimental solubility H₂S-CO₂-C_nH_2n-H₂O-salt data performed above 10,000 psig.^29,38 

3. Figure 10 illustrates the complex interactions between fugacity, activity, and dissolved concentration predicted at 75°F (24°C) and 5,000 psig (34.5 MPa). In this example, as the salinity increases from 0 to 150,000 mg/L Cl⁻, the activity coefficient (γ_H2S) increases from 1 to 1.6, while C_H2S decreases by nearly ½. Conversely, ϕ_H2S remains constant (∼0.3) and thus insensitive to changes in ionic strength. Under the conditions described, f_H2S is fixed, which implies a_H2S is also fixed. While C_H2S and γ_H2S vary inversely, when the two parameters are multiplied together, they should reproduce the fixed a_H2S, and thus should parallel f_H2S, which is also fixed. In the end, the aqueous activity (as defined here) together with the Poynting correction should deliver the same result as equating a known gas f_H2S with the aqueous f_H2S. It is worth noting that, in the context of SSC, salt content not only has an effect on H₂S, but it also affects other aspects of the mechanism (e.g., in situ pH and film passivation of the metal oxide).⁽⁴⁾

   FIGURE 10.

   

   View largeDownload slide

   Model calculations⁽³⁾ performed at a water gas ratio (WGR) of 10 bbl/MMscf, 75°F (24°C), 5,000 psig, and pH 3. The non-aqueous phase contains 60 mol ppm H₂S, 10 mol% CO₂ and balance CH₄. Shows the change in γ_H2S, φ_H2S, and C_H2S as a function of [Cl⁻].

   FIGURE 10.

   

   View largeDownload slide

   Model calculations⁽³⁾ performed at a water gas ratio (WGR) of 10 bbl/MMscf, 75°F (24°C), 5,000 psig, and pH 3. The non-aqueous phase contains 60 mol ppm H₂S, 10 mol% CO₂ and balance CH₄. Shows the change in γ_H2S, φ_H2S, and C_H2S as a function of [Cl⁻].

   Close modal
4. To calculate the thermodynamic parameters using current ionic-EOS frameworks in the range of 10,000 psig to 20,000 psig (69 MPa to 138 MPa) total pressure, requires significant extrapolation of the datasets to capture the C_H2S typical of mildly sour HPHT wells. The concern is the experimental H₂S solubility data collected to date were: (1) conducted at total pressures that are nearly half an order of magnitude lower than those expected for HPHT wells and (2) the P_H2S used to calculate H₂S solubility are orders of magnitude higher than those typical for mildly sour service. The technology gap between the experimental data and HPHT domain of interest is shown graphically in Figure 11.

   FIGURE 11.

   

   View largeDownload slide

   Schematic comparison between the available H₂S-H₂O-salt public database domain versus the HPHT sour service domain of interest.

   FIGURE 11.

   

   View largeDownload slide

   Schematic comparison between the available H₂S-H₂O-salt public database domain versus the HPHT sour service domain of interest.

   Close modal

In summary, there is insufficient data available to properly verify the accuracy of the existing ionic-EOS frameworks in environments reflective of mildly sour HPHT wells, because H₂S is toxic (which limits the number of facilities able to conduct solubility research) and simulating HPHT conditions in the lab with methane as the dominant gaseous constituent is difficult. With a few exceptions, the motivation of thermodynamicists and software developers up to this point has been to primarily develop a “universal” ionic-EOS tool, not a dedicated HPHT-mildly sour service model.

Extension of the existing H₂S-H₂O-salt experimental database to the HPHT domain of interest is critical, but just knowing the relationship between P_H2S and C_H2S at HPHT conditions is not enough. To be of practical use to the industry, C_H2S at full wellbore pressure must also be tied back to the material-specific SSC threshold, usually defined as the P_H2S limit at normal temperature and pressure (NTP) as implied in Figure 4 for example.

A promising method to correlate HPHT sour phase behavior to low-pressure SSC tests^5-6  is an empirical approach that measures C_H2S. The direct approach assumes that C_H2S is the parameter that influences the SSC mechanism. The dissolved H₂S-SSC correlation metric rearranges the Ensemble Henry’s Law so that all thermodynamic parameters are grouped together (Equation [5]):

This is simplified to Equation (6)

where ^Rk_H2S represents the real (apparent) H₂S Henry’s Law coefficient. The coefficient, ^Rk_H2S, has the units of mg/L/psia,⁽⁵⁾ when P_H2S is in units of psia, and C_H2S has the dimension of mg/L.^11,20,40 Rk_H2S is temperature, total pressure, hydrocarbon composition, and ionic strength dependent due to nonideal solubility effects.⁴⁴  At NPT and no salt, ^Rk_H2S equals ^Ik_H2S. As total pressure increases, ^Rk_H2S decreases logarithmically with increasing total pressure; as P_T approaches 20,000 psig, ^Rk_H2S << ^Ik_H2S.⁴⁵ 

Critics¹³  of this approach argue that by defining sour severity based on C_H2S there is the potential for higher risk because the over-conservative safety factors intrinsic to the P_H2S and f_H2S approaches are removed. Proponents of C_H2S approach^5-7,28,46  have argued that knowing the sour severity limit does not restrict the user from setting their own safety factor when qualifying materials. Typically, risk-based design tolerances are often component (e.g., primary vs. secondary pressure containment) and well (e.g., onshore vs. offshore) specific. Utilization of a dissolved H₂S design approach, however, does require accurate analytical measurements and a robust experimental database to fully validate the methodology.

Before this dissolved H₂S approach could be formally adopted, the effect of the C_H2S on the SSC mechanism needs to be quantified and properly characterized. Because the C_H2S is a requirement for SSC, the authors argue that the same parameter may also be the characterizing parameter for SSC. Evidence of a correlation between C_H2S and fracture toughness as a function of total pressure have already been demonstrated.^10,14  In the absence of additional experimental data, fugacity remains the more conservative approach.^8,10,13  Resorting to a C_H2S-SSC correlation metric can only be justified by understanding its role in the SSC mechanism, not just by measuring concentrations.⁽⁷⁾

In summary, despite the many challenges, the benefits of the C_H2S design approach are substantial, and this approach is already being recommended for low-cost, high-pressure wells, with the caveat that accurate H₂S measurements in the field must be obtained.⁷ 

A hypothetical sour service example is used to demonstrate the utility of the dissolved H₂S approach. In the following example, a producer well has a water gas ratio (WGR) of 10 bbl water per 1 MMscf gas. The hydrocarbon phase is composed of 60 ppm H₂S + 10 mol% CO₂ + balance CH₄ and the gas is assumed to be comingled with freshwater containing 0 mg/L Cl⁻. Modern ionic-EOS calculations⁽³⁾ predict the following sour service metrics at 20,000 psig and 75°F (24°C): 1 psia (6.9 kPa) P_H2S (hydrocarbon) ≡ 0.32 psia (2.2 kPa) f_H2S ≡ 12 mg/L H₂S (aqueous). The implications of each sour severity metric are discussed below.

If a 20,000 psig autoclave is available for material qualification, then regardless of the sour severity metric selected, the same SSC outcome is expected. As mentioned earlier, there are several logistical problems related to utilizing autoclaves rated to 20,000 psig. As such, users often prefer using low-pressure rated equipment to qualify materials. Designing an SSC qualification test with an exposure gas of P_H2S = 1 psia at P_T = 1 atm may place a severe restriction on available material options for users.⁽⁷⁾

Alternatively, candidate materials could be qualified at f_H2S  = 0.32 psia.⁽⁷⁾ When designing an SSC material qualification test using low-pressure rated equipment and matching lab sour severity based on equivalent f_H2S predicted for the HP field environment, the following factors should be considered: (1) the ϕ_H2S varies with total pressure⁹  (refer to Figure 8[a]), (2) the C_H2S also changes with total pressure and its behavior is independent of ϕ_H2S, and (3) a single f_H2S value can correspond to multiple C_H2S, which may lead to different cracking outcomes. Figure 13 shows the change in C_H2S with total pressure at three different y_H2S ranging from 60 ppm to 240 ppm. As the total pressure exceeds the bubble point pressure, the C_H2S levels off. At 18,000 psig (124 MPa) and at 60 ppm yH₂S, f_H2S = 0.32 psia and C_H2S = 12 mg/L. This same f_H2S = 0.32 psia value also corresponds to C_H2S = 23 mg/L at P_T = 11,300 psia (78 MPa; y_H2S = 120) and C_H2S = 45 mg/L at P_T = 4,700 psig (32.4 MPa; y_H2S = 240 ppm).

A single fugacity value corresponding to multiple C_H2S is of concern as the Standard¹  assigns material-specific P_H2S limits indiscriminate of total pressure. While a single C_H2S can also correspond to multiple f_H2S values, mathematically the inverse operation is noncommutative as SSC is ultimately dependent upon the H₂S content of the aqueous phase. To be meaningful, the f_H2S threshold must be reported with the specific PVT reference state used to make the thermodynamic calculations and to correctly design material qualification tests in the laboratory.

One reservation against adopting a dissolved H₂S-based sour service metric is C_H2S is meaningless to most users. To make 12 mg/L C_H2S meaningful, and to correctly interpret SSC qualification test results that are often evaluated using low-pressure rated equipment, the C_H2S expected under HPHT conditions ought to be translated to the equivalent P_H2S at NTP.

Based on the above observations, the authors refer to this as the C_H2S-SSC Scalability Paradigm:

An important limitation to note is Figure 17 does not account for the effect of temperature and could pose problems when translating from a HP field to an equivalent lab condition. Because SSC is a low-temperature phenomenon, however, the temperature range of interest is narrow. Based on the calculations shown in Figure 8(c), between 77°F and 200°F (25°C to 93°C) the predicted C_H2S barely changes for a fixed y_H2S. Thus, the concern is not at high temperatures, but rather applying the proposed diagram to assess SSC risks at temperatures <70°F (21°C). Gas solubility is known to increase with decreasing temperature. The resolution is to generate an equivalent diagram at 40°F (4°C), which is generally accepted as the lower service temperature of an offshore well at the seafloor.

Much work is needed before submitting a formal Ballot to the NACE MR0175/ISO 15156 Maintenance Panel to revise the historically accepted definition of “environmental severity” based on P_H2S to one based on C_H2S. A concerted effort to experimentally verify the behavior of H₂S under HPHT conditions relevant to the oilfield is underway.⁴⁵  Verification of the C_H2S-SSC scalability paradigm requires the use of high-pressure equipment up to 20,000 psig, and H₂S toxicity concerns must be handled with care. Accurate analytical documentation of the exposure environment must be provided to irrefutably link C_H2S to the cracking behavior of high-strength materials at multiple total pressures. For this benchmark study, materials with controlled chemistries, microstructures, and mechanical properties are recommended.

While the challenges to validating the C_H2S-SSC scalability paradigm are steep, they are not insurmountable, and the potential rewards are much higher. In fact, there is already reliable research on this important topic. In 2018, Chambers, et al.,¹⁴  exposed T95 carbon steel in multiple continuous-liquid-loaded autoclaves, which were elegantly designed to deliver equal C_H2S at increasing total pressures (from 14.7 psig to 10,000 psig) for 30 d. The fracture toughness of the material reportedly increased with increasing total pressure beyond statistical uncertainty, although the results were not expected and the authors noted several experimental complications.¹⁴  Further investigation is warranted to confirm the test results independently, and to possibly extend the application to additional metallurgies.

Having thoroughly described a process to validate the C_H2S-SSC scalability paradigm, our plan is to spearhead an ambitious multi-year testing program to prove our hypothesis. The benefit to operators is the justification to safely select less expensive materials, traditionally considered marginal, as suitable candidates for mildly sour HPHT environments without the risk of failure by SSC. For material manufacturers, validating the paradigm will expand the application of existing products to new sour service applications. In conclusion, confirmation of the paradigm will reduce material qualification testing costs for most applications as low-pressure rated equipment (e.g., <5,000 psig) could be used to reliably predict SSC behavior under HPHT conditions (i.e., 10,000 psig to 20,000 psig).

As accurate C_H2S and y_H2S measurements alone are insufficient to demonstrate validity of this approach to characterize severity in the specific SSC context, the authors have recently performed a limited number of SSC tests with 15Cr-125 martensitic stainless steel (MSS) in batch-loaded autoclaves up to 20,000 psig and 70±5°F (21±3°C). To investigate the causal relationship between C_H2S and SSC, all environmental conditions were held constant (including a fixed molar inventory of H₂S) per autoclave test, with exception of increasing the total pressure with additional N₂ booster gas. To date, the unpublished results are consistent with our C_H2S-SSC scalability paradigm: i.e., the severity of the physical damage induced by SSC (in terms of both cracking and pitting features) are closely related to the measured C_H2S and independent of the measured y_H2S/pseudo-P_H2S at increasing P_T. We plan to present our findings in a future publication.

On a final note, it is recognized that capitalizing on all of these advantages cannot be done cavalierly. While there are obvious “capital expenditure” advantages in terms of more affordable materials if SSC is controlled by C_H2S, caution needs to be applied in confirming this fact and broad applicability must be demonstrated before the “cost” advantages are fully realized.

• While P_H2S may be appropriate for assessing sour severity for low-pressure wells, it is not a predictive parameter within the HPHT domain, because the relationship between P_H2S and C_H2S deviates from simple proportionality. Defining sour service qualification parameters based on P_H2S for materials intended for HPHT application generally leads to over-conservatism.

• ISO 15156-1:2020 now permits users the option of implementing either the P_H2S, f_H2S, a_H2S, or C_H2S as metrics to define environmental severity. Of the metrics recognized by the standard, C_H2S is the only modern metric that is directly experimentally measurable.

• A C_H2S-SSC paradigm could be used to translate historical SSC material testing experience (generally obtained at total pressures <5,000 psig [34.5 MPa]) to predict SSC behavior between 10,000 psig and 20,000 psig (69 MPa to 138 MPa). Such a strategy could also be adopted for other modern sour severity metrics as well.

• While both f_H2S and C_H2S design metrics can be used to predict sour severity under HPHT conditions, thermodynamic predictions should be verified with physical testing. In the absence of additional experimental data, however, f_H2S remains the more conservative approach, pursuant the thermodynamic calculations are performed correctly using a validated ionic-EOS framework.

• Solving the Ensemble Henry’s Law equation for C_H2S as written requires the use of sophisticated ionic-EOS frameworks. However, extending the use of these frameworks to the mildly sour and high total pressures requires significant extrapolation (with concomitant inaccuracies) of currently available H₂S-H₂O-salt databases.

• When thermodynamic correction factors are combined with the Strict or Ideal Henry’s Law constant into a single H₂S solubility term, the resulting ^Rk_H2S can be experimentally confirmed. As a secondary benefit, if ^Rk_H2S values are experimentally confirmed, this implies confirmation of the fugacity EOS models as well.

• For a given sour gas composition, the predicted relationship between C_H2S and P_T showed C_H2S remained nearly constant from approximately 1,800 psig P_T out to 20,000 psig, rendering C_H2S as a potential scalable H₂S-SSC parameter candidate.

• In the absence of experimental confirmation, however, caution needs to be applied when considering the implications of applying the dissolved H₂S approach.

The authors wish to thank Honeywell Corrosion Solutions (Houston, TX) for its support of time and underwriting this study. A.J. Gerbino, OLI Systems, Inc., is gratefully acknowledged by the authors for his guidance and continued support of the research. The authors would also like to acknowledge the thorough analysis and thoughtful critique provided by the reviewers.

a_i

Activity of i in the aqueous phase, mol/L

C_i

Dissolved i concentration in the aqueous phase, mg/L

f_i

Fugacity of i in the gas phase, psia

i

Component

k_i

Henry’s constant of i solubility, mg/(L·psia)

n_i

Number of i moles in gas phase, mol

P_i

Partial pressure of i, psia

P_T

Total pressure, psig

R

Ideal gas constant, 8.3144 J/(mol·K)

T

System temperature, K or °F

V

Volume of the gas phase, L

x_i

Mole fraction i in the aqueous phase, mol%

y_i

Mole fraction of i in gas/liquid hydrocarbon phase, mol% (mol ppm)

γ_i

Activity coefficient of i in the aqueous phase—Corrects for the nonideal behavior of i in the aqueous phase

ξ_i

The Poynting correction factor is often expressed as an exponential term related the partial molar volume of i. It adjusts for the effect of total system pressure on the solubility of i in the aqueous phase

μ_i

Chemical potential of i, J/mol—fundamental parameter derived from Gibbs free energy

∅_i

Fugacity coefficient of i in the gas phase—a correction for nonideal behavior of H₂S in the hydrocarbon phase