PREVENTING HTHA FAILURES WITH

THE BUCKEYE MODEL OPERATE SAFELY IN A

HYDROGEN ENVIRONMENT

HTHA DAMAGE

High-Temperature Hydrogen Attack (HTHA) is the progressive degradation of carbon and low-alloy steels
exposed to hydrogen at elevated temperatures. Accurately predicting HTHA failures has proven to be one of
the most elusive problems challenging the industry for more than half a century. Although basic guidelines
for choosing ostensibly safe combinations of temperature and hydrogen partial pressure are provided in API
RP 941, there continue to be unexpected failures. This prompted the development of a more sophisticated
approach for the prediction of HTHA remaining life.

E²G HTHA JOINT INDUSTRY PROJECT

In 2013, E²G initiated an ongoing Joint Industry Project (JIP) to focus on three essential tasks necessary to better
understand HTHA life prediction:

Develop a predictive model of Evaluate and improve the ability Perform additional testing in
HTHA damage progression and of current NDE methods to detect hydrogen environments to further
ultimate failure. existing HTHA damage. improve failure predictions.

» This effort has led to the most powerful tool for HTHA life prediction to date – the E²G Buckeye Model.

HTHA DAMAGE PROGRESSION OVER TIME

KEY
Damage Parameter

New Material
~10 Years in Service
~20 Years in Service
Time at Failure

Thickness Profile

Damage prediction of a carbon steel heat exchanger operating at ~750° F at ~200 psia hydrogen

PREDICTIONS YOU CAN TRUST

The true test of any model is how well it predicts real-world failures. Well-documented data on HTHA failures
is limited; however, the Buckeye Model has been shown to be consistent with reliable failure data from API RP
941 and other sources.
THE E²G BUCKEYE MODEL NEW FITNESS-FOR-SERVICE PROCEDURES

E²G’s Buckeye Model integrates the primary physical The Buckeye Model forms the basis for a new set of
mechanisms known to be responsible for HTHA damage: Fitness-for-Service (FFS) rules that have been delivered as
diffusion of hydrogen through metal, thermodynamics and part of the JIP work conducted by E²G. The remaining life is
kinetics of methane formation, growth of methane-filled predicted separately for two modes of failure:
voids by creep of the surrounding material, and the increased
tendency for cracks to initiate and grow. By integrating these
1 2
mechanisms into one cohesive model, E²G has developed
the first truly comprehensive and practical method of HTHA Volumetric failure when Crack growth failure
damage prediction. the distributed damage when the assessment
(void density) reaches a point lies outside of
The most important aspect of the model is that it accounts critical level. the Failure Assessment
for the time-dependent nature of HTHA. The Nelson Curves Diagram (FAD).
in API RP 941 do not account for this time dependence.
Additionally, the model accounts for other important factors
such as the following: As with most FFS procedures, increasing levels of sophistication
are employed. A Level 1 FFS assessment uses a set of graphical
» Variable operating conditions screening curves that resemble the traditional Nelson Curves,
» Type and thickness of material (based metal and cladding) except they also account for the time-dependence of HTHA.
» Heat treatment (PWHT vs. non-PWHT) These time-dependent curves are named the Prager Curves in
» Welding residual stress honor of Dr. Martin Prager (Director of The Welding Research
» HTHA crack-like flaws analysis Council, Inc.), the originator of this approach and a close
» Existing damage (including localized damage) collaborator with E²G on the HTHA JIP project.

TIME-DE P E N DE N T P R AG E R C U RV E S EMBRACING UNCERTAINTY

LEVEL 1 PRAGER CURVES PROBABILITY OF FAILURE

100%
90%
80%
Temperature

70%
Increasing Hours in Service
60%
50%
40%
30%
20%
Hydrogen Pressure 10%
0%
0 20 40 60 80 100 120
Age (Years)

Failing a Level 1 FFS assessment is not the end of the story. Our experts at E²G assess components using the Buckeye
Instead, a more sophisticated Level 2 FFS can be performed by Model. Detailed information for equipment operating in HTHA
running the Buckeye Model for HTHA assessment. environments is often limited. Statistical methods are used to
characterize uncertainty, and probabilistic models are used to
estimate the probability of failure over time.

CORPORATE HEADQUARTERS SATELLITE OFFICE CONTACT

20600 Chagrin Boulevard, Suite 1200 Houston, TX 216.283.9519
Shaker Heights, OH 44122 Sales@E2G.com
www.E2G.com