The Coating That Cracked 36 Bolts: What a Gulf of Mexico LMRP Separation Teaches About Fastener Finishes

#TLDR

On December 18, 2012, the lower marine riser package (LMRP) on the Transocean Discoverer India separated from the blowout preventer stack at Keathley Canyon block KC-736 in the Gulf of Mexico. About 432 barrels of synthetic-based drilling fluid spilled into the water. Chevron reported the cause to the Bureau of Safety and Environmental Enforcement (BSEE): failure of the GE Oil & Gas (formerly Vetco-Gray) H4 connector bolts that held the connector halves together. BSEE’s QC-FIT Evaluation of Connector and Bolt Failures (August 2014) and the industry root-cause work it reviewed agree on the primary mechanism: hydrogen-induced stress corrosion cracking from hydrogen embrittlement. A decisive process error sat upstream of the water. A third-tier coating vendor electroplated the bolts to an outdated ASTM B633 edition and skipped the post-electroplating bake the 2007 standard required for steel at that strength. The finish was supposed to protect the bolts. Done wrong, it helped load them with hydrogen before they ever left the shop.

What happened on the seafloor

The H4 bolts were serious hardware: roughly 9 inches long, about 2 inches in diameter, AISI 4340 alloy steel at Class 145 yield strength (about 145 ksi), hardness specified between 34 and 38 HRC. Thirty-six of them fastened the LMRP connector. On December 18, 2012, all 36 failed. The LMRP lifted off the BOP stack. Synthetic mud vented into the Gulf.

GE issued Safety Notice SN 13-001, expanded the recall globally, and shipped more than 10,000 replacement bolts for 361 LMRP connectors worldwide. BSEE issued Safety Alert No. 303 and stood up the Quality Control–Failure Incident Team (QC-FIT) to ask a harder question: was this one bad lot, or a systemic problem in how the industry designs, coats, and qualifies subsea fasteners?

What went wrong with the fastener “selection”

The steel grade was intentional. High hardness and yield were specified so the connector could carry tensile, bending, and axial loads in service. The coating choice was intentional too. Before 2007, H4 bolts used a zinc phosphate finish. After 2007, the finish changed to zinc chromate electroplating under ASTM B633 Type II, service class SC 2, for better saltwater corrosion resistance. Sacrificial zinc is a common way to protect carbon and alloy steel.

Hydrogen is the catch. Electroplating introduces atomic hydrogen into high-strength steel. Industry practice and modern coating standards answer that risk with controlled baking: heat the parts after plating so trapped hydrogen can diffuse out before the bolt goes into tension in the field. The 2013 combined root-cause analysis by Chevron, Transocean, and GE, and BSEE’s QC-FIT report, both state that a GE subcontractor relied on the 1998 edition of ASTM B633 instead of the 2007 edition. The older edition did not require post-bake at the strength level used for these H4 bolts. The bolts received a pre-bake. They did not receive the post-bake the 2007 standard required for hardness above about 31 HRC or tensile strength above about 145 ksi.

That is a fastener-system failure as much as a materials failure. The “right” alloy with the “wrong” finish process is still the wrong fastener for the job. GE’s quality system at the time audited first-tier suppliers, not the third-tier plater. Neither the OEM’s QMS nor the operator and contractor mechanical-integrity programs caught a multi-year gap against the current coating standard. BSEE also flagged missing paint on bolt-head faces toward the connector body as a possible contributor: bare steel under cathodic protection can draw more current and generate more hydrogen at the surface. QC-FIT further noted that hardness above roughly 34 HRC remains a known concern for hydrogen damage in subsea service, and that industry standards still disagreed on maximum hardness for marine bolting.

The correct coating and process choice

Match the finish to the steel and the service:

1. If you electroplate high-strength steel, bake it correctly. Follow the current edition of the coating standard that applies (for zinc electrodeposits, ASTM B633 and related guides such as ASTM B850 on post-coating treatments). For hardness and strength in the H4 range, that means both the pre-plate and post-plate thermal cycles the standard requires. “We plated it” is not the same as “we made it safe.”

2. Prefer finishes with a known hydrogen story for high-hardness bolts. GE’s response to the incident included reverting replacement H4 bolts to the earlier zinc phosphate system with a specified post-bake, citing no prior failures with that system on the same stack’s lower connector. Phosphate is not a miracle coating. The lesson is process control and fitness for hydrogen-sensitive steel, not brand loyalty to any one finish name.

3. Question coating class and environment. QC-FIT read ASTM B633’s SC 2 examples as moderate, mostly dry, occasional-condensation service (tools, machine parts), and questioned whether SC 2 thickness and duty class fit continuous subsea exposure. GE disagreed and pointed to assembly thickness limits in API documents. When experts read the same standard two ways, treat that as a design review item: confirm service class, thickness, and marine suitability in writing before you freeze a bill of materials.

4. Control hardness where hydrogen and cathodic protection meet. Several marine and sour-service references (including guidance cited by QC-FIT from NACE/ISO and NORSOK families) push maximum hardness lower than 38 HRC for many subsea duties. Strength and toughness are a trade. Blindly maximizing HRC in seawater under CP is how you buy embrittlement risk.

5. Qualify the whole supply chain. A fastener is not “approved” because the first-tier vendor is on an approved list. Plating, heat treat, and raw stock sit where hydrogen and hardness are set. Audit the tiers that touch those steps, and lock the edition of every ASTM/API/ISO document on the purchase order.

Load, environment, application

Subsea connector bolts live at the intersection of high preload, cyclic and shock loads (including jarring), seawater chlorides, and cathodic protection that can feed hydrogen to bare steel. Bay Bridge taught a related hydrogen lesson on land with galvanized A354 BD rods under sustained tension. This Gulf of Mexico case teaches the manufacturing twin: the coating process can charge the bolt with hydrogen before installation, then service tension and the marine environment finish the crack.

The right fastener for this job is not “the strongest 4340 stud you can buy.” It is a stud whose grade, hardness window, coating chemistry, bake cycle, paint/CP interfaces, and supplier controls all match deepwater connector duty. Skip any one of those, and thirty-six correct-looking heads can still part a riser package.

Closing

BSEE’s public record is blunt: hydrogen-induced SCC from embrittlement, driven in large part by electroplating without the post-bake the then-current ASTM standard required, on high-hardness alloy bolts in a marine connector. The industry replaced tens of thousands of bolts and reopened standards work on subsea bolting. The shop-floor takeaway is simpler. Coatings are structural decisions. Finish, bake, hardness, and environment are part of fastener selection, not paperwork after the steel is chosen. Put the right fastener, with the right process history, on the job, or the job will teach the lesson the hard way.