Stress Corrosion Cracking of Frac Pump Fluid Ends

[ttocsewol]

We've seen numerous fluid end failures due to cracking, and have several engineering reports in hand identifying stress corrosion cracking(SSC), or corrosion-fatigue cracking as the reason for failure.

The fluid ends are constructed of an alloy steel, and are pumping water with many similarities to seawater. Failures are occuring at 400 to 600 hours of duty. The duty consists of 20 hours per day pumping at very high rate and 6 or 7 thousand psi. The cracks are sporadic in where they occur, but all appear to originate at a corrosion pit.

Prior to getting the engineering reports, we were looking at fatigue failures due to surge and resonance, and while these cannot be eliminated, are likely contributors to the early failure.

Several questions I have:
Could the pitting corrosion being seen simply be electrochemsitry at work rather than an environmental cause?

Would the pitting corrosion we're seeing in the fluid ends be reduced by implementing some cathodic protection?

 

[btrueblood]

"an alloy steel" is kinda vague. Could you pin it down a bit more?

 

[ttocsewol]

Intentionally vague, btrue. The alloy was designated by the manufacturer as 4S16854, but that means nothing to me, and I'm not sure would mean anything to anyone else.

One analysis has it as: 0.3% C, 0.5% Mn, 1.1% Cr, 0.5% Mo, 2.6% Ni, 0.2% Cu, 0.1% V, and 0.2% Si

This doesn't appear to be a standard alloy, although I'm no metallurgist.

 

[stanweld]

The alloy is essentially a derivation of AISI 4330 with more nickel. I would assume that it has been heat treated and is quite hard (tempered martensitic microstructure). I would readily suspect pitting corrosion in the suggested environment. Pitting corrosion can evolve hydrogen, which can lead to SCC.

 

[stanweld]

I should also add that the 4330 derivation also has added Vanadium.

 

[EdStainless]

and/or the pits cause local stress concentration.
Corrosion will be accelerated down in the pits because of all of the regular reasons.
Could you even do CP in this application? It would help.

I believe that some coatings have been tried in such applications.

= = = = = = = = = = = = = = = = = = = =
Plymouth Tube

 

[ttocsewol]

Thanks for the responses so far.

Stanweld, you've assumed correctly. The heads are heat treated to about 140K psi yield strength. They are also autofrettage treated, which has been indicated on one report as being unnecessary, and may potentially exacerbate the situation.

The reason for our quandry is the failure rate is atypical, it is uncommon for fluid ends to fail at less than 1000 hours, and many of ours that have seen different duty with comparable water have more than 2000 hours on them. The high failure rate can be traced to one type of use for one customer, and appears to be an environmental problem.

EdStainless, we haven't investigated the coatings possibility, although we have looked at an ultrasonic treatment that I don't have much information on, but has taken a significant investment.

Any further input would be welcome. I've included a pic this time.

 

[metengr]

The picture does confirm a hydrogen-assisted SCC mechanism in service because of the rather straight profile of the cracks and families of cracks being observed. I believe a barrier coating like epoxy should be investigated provided it is rated for this type of service environment.

The cracking will re-occur in this material because of the specified high strength heat treatment condition. Is there a reason for having this high strength alloy? What about nodular iron in this application?

 

[TVP]

I agree with metengr, definitely SCC. This alloy and heat treatment condition will be susceptible to SCC in a seawater environment. Options include the following:

- Cathodic Protection
- Lower strength with the same alloy (300 HB or 32 HRC maximum hardness, UTS < 105 ksi)
- Coating

 

[SJones]

CP will generate hydrogen - stops the pitting, but not the cracking in this high strength material if the stress raisers are present.

Steve Jones
Materials & Corrosion Engineer

http://www.linkedin.com/pub/8/83b/b04

 

[stanweld]

I would suspect that the one customer uses a more corrosive fracking fluid (tweaked the formulation just a tad) than the other customer(s).

 

[ttocsewol]

Presumably, the reason for the high strength heat treatment is to achieve a higher pressure rating. The ends and the iron tubulars following it are rated to 15k psi. We run at 50 to 60% rated capacity.

Metengr, the "hydrogen-assisted SCC mechanism" you spoke of, does this require sulfide production? Several factors that may relevant to that, the fracking fluid consists of 99.9% water, however the water is reused many times, is laden with bacteria, is non-aerated and so is greatly oxygen-deficient, is at 70 to 100F in temp, and is minimally treated with a biocide seconds before entering the pumps.

I have checked for sulfides, and if present are below our detection limit. SRB, IRB, and APB are all present in the water, but it appears that the biocide keeps them from proliferating.

We think that part of the decline in fluid end longevity iss the constant immersion for 20 hours a day. In the past, our jobs were an hour or two, and within 5 hours, the pumps were drained. We also typically saw better water quality.

The epoxy coating may deserve some consideration. It would need to be applied to only non-contact parts of the fluid end, but this is where most of the pitting corrosion is occurring. Occasionally, sand proppant is run in the frac fluid, so it would have a reduced life, but may buy us some time.

 

[stanweld]

I suspect that the required hardness is to extend service life when pumping in sand to prevent erosion. I doubt that an epoxy type coating would be effective.