Hydrogen sulfide is a known reliability accelerant for rotating equipment. H2S-containing process streams — common in sour gas gathering, NGL fractionation with high-sulfur crude input, and amine treating unit recirculation loops — degrade equipment through several mechanisms that standard vibration baselines don't account for. A condition monitoring program that applies the same alarm setpoints and rebaseline intervals to a sour gas centrifugal compressor as it does to a sweet gas unit is operating with a false model of the equipment's degradation rate. This article covers the specific mechanisms, their monitoring implications, and the practical adjustments required to maintain detection reliability in sour service.
Sulfide Stress Cracking and Its Effect on Rotating Equipment Metallurgy
Sulfide stress cracking (SSC) is an environmentally assisted cracking mechanism that occurs when susceptible high-strength steel components are simultaneously exposed to H2S and under tensile stress. The mechanism is governed by NACE MR0175 / ISO 15156, which specifies material hardness limits, heat treatment requirements, and testing protocols for equipment in H2S service. The standard defines "sour service" conditions as those with a partial pressure of H2S exceeding 0.0003 MPa absolute (0.05 psia) in gas or two-phase systems.
For rotating equipment, the components at highest risk from SSC are high-strength steel shafting, compressor impellers and thrust collars in non-austenitic materials, fasteners, and the case-hardened surfaces of rolling element bearings. API 617 (centrifugal compressors) and API 618 (reciprocating compressors) both reference NACE MR0175 material requirements when specifying compressor components for sour service. When the metallurgy is correct per NACE, direct SSC failure of rotating components is relatively rare; the more common problem is that incipient surface damage — micro-cracking in a bearing raceway or impeller blade root — propagates faster in an H2S environment than in clean service, compressing the degradation timeline.
Accelerated Bearing Degradation Mechanisms in H2S Service
Three distinct mechanisms drive accelerated bearing degradation in sour service environments:
Hydrogen embrittlement: Atomic hydrogen produced by electrochemical reactions at H2S-exposed metal surfaces can diffuse into bearing steel and embrittle the material, reducing resistance to fatigue crack propagation. This lowers the effective L10 bearing life below what the bearing manufacturer's rating would predict for the applied load and speed. A bearing rated for 50,000 hours at a given load in clean service may deliver 30,000–40,000 hours in a moderately sour environment — the factor varies with H2S concentration, temperature, and actual bearing stress.
Corrosion of bearing surfaces: In wet sour gas service (where liquid water is present with H2S), active corrosion of bearing raceways can occur during extended idle periods. Bearings that sit idle in a wet H2S atmosphere during planned or unplanned shutdowns accumulate pitting damage that manifests as elevated broadband noise and early spall formation when the machine returns to service. This post-startup degradation pattern is diagnostically significant: a bearing that was monitoring clean at shutdown may show elevated kurtosis and acceleration within hours of restart.
Seal face material degradation: Mechanical seal faces in sour service use different materials (typically silicon carbide against carbon with special face coatings) than sweet service seals. When seal barrier fluid systems are not maintained — barrier fluid contamination, low barrier pressure, pump-out ring degradation — H2S can reach the seal faces and attack the carbon face, producing face wear rates 2–5 times higher than clean service. This is not a vibration-detectable failure mode in its early stages; it manifests as abnormal barrier fluid consumption and seal pot pressure behavior first, and as vibration changes (1× amplitude from face imbalance) only in advanced failure.
Adjusted Monitoring Parameters for Sour Service Equipment
The monitoring adjustments for sour service equipment follow logically from the degradation mechanisms:
Tighter alarm thresholds: Because H2S service compresses the degradation timeline, a 25% increase above baseline vibration that provides 3–4 weeks of lead time in sweet service may provide only 1–2 weeks of lead time in a sour application. Setting alert thresholds at 15–20% above baseline (rather than 25–30%) for sour service equipment preserves equivalent maintenance planning lead time. This requires accepting a marginally higher false-alert rate for process upsets, which should be managed by correlating alerts with process variable context (H2S concentration spikes, startup events) rather than by widening the threshold back to sweet service levels.
Shorter rebaseline intervals: Equipment baseline vibration can shift over time as the machine runs in normal operation — this is acceptable and expected for long-run equipment. In sweet service, rebaseline intervals of 6–12 months are typical. For sour service equipment, a 3–6 month rebaseline interval is more appropriate, because the gradual deterioration in bearing L10 life means that "normal" for a 2-year-old bearing in sour service is genuinely different from "normal" for that bearing when it was new.
Post-startup monitoring emphasis: Given the corrosion-during-idle mechanism, sour service bearings deserve elevated monitoring attention in the first 24–72 hours after startup. A bearing that shows kurtosis escalation (from a clean 2.5 to 6+ in the first hour of operation) following a restart from a wet idle period is showing corrosion-damage symptoms, not a new installation run-in pattern.
NEC Class I Division Requirements and Sensor Selection
Facilities handling H2S streams are classified as hazardous areas under the National Electrical Code. NEC Class I Division 1 (or Division 2, depending on concentration probability and ventilation) designations apply to compressor areas and pump stations handling sour gas. This affects sensor selection for condition monitoring: vibration transmitters, proximity probe drivers, and wireless devices installed in Class I Div 1/2 areas must carry appropriate hazardous area approvals (UL Listed / FM Approved for the applicable division, or ATEX/IECEx for international installations).
Bently Nevada 3300 and 3500 series transducer systems have documented hazardous area certifications. Third-party accelerometer transmitters vary widely in their certification status. When specifying condition monitoring sensors for addition to existing sour gas equipment, verifying hazardous area certification is a non-negotiable prerequisite — not a commissioning afterthought.
We are not saying that every sensor approved for general industrial use is acceptable in a Class I Div 1 area — it is not. We are saying that the certification verification step is often missed in early-stage condition monitoring deployments where sensors are selected based on technical specifications alone, without confirming area classification compatibility.
Practical Baseline Management for Mixed Sweet/Sour Fleets
A growing midstream company operating both sweet and sour gas equipment — common in areas like the Permian Basin where different formations produce gas at different H2S levels — faces the challenge of managing two different monitoring regimes within the same facility or fleet. The simplest implementation is to tag each asset in the condition monitoring system with its service classification (sweet / sour mild / sour severe) and apply class-specific alert threshold multipliers, rebaseline intervals, and post-startup monitoring flags automatically.
This service classification should be derived from design conditions, not current process readings. An amine absorber recirculation pump designed for sour service may be operating on a sweet inlet during a low-H2S production period, but its metallurgy, seal type, and bearing specification are still sour-service design — and the monitoring parameters should reflect that, because operating conditions can change without notice when upstream production changes composition.
The OREDA reliability database and ISO 14224 equipment taxonomy both recognize sour service as a distinct reliability category with statistically different failure rates than sweet service. Using these references to calibrate expected MTBF for sour equipment — and tracking actual MTBF performance against those benchmarks — provides a closed-loop check on whether the monitoring program is delivering the maintenance planning lead time it was designed to provide.
Midstreamly supports service-class-tagged threshold configuration for sour and sweet equipment in the same fleet. Talk to our engineering team about your sour gas monitoring requirements.
