Heat Exchanger Material

[metengr]

For those of you MET Engr's that deal with power plant heat exchangers, I came across an interesting problem with 304N tube material in a certain manufacturer's high pressure (HP), horizontal, U-bend design feedwater heater. The problem was OD intiated cracks in the 304N feedwater heater tube material located within the desuperheater zone of our HP heaters. The cracks were circumferential in orientation. Metallurgical analysis of several tube samples confirmed fatigue crack propagation(transgranular)with some crack branching midwall and near the crack tips (similar to TGSCC). We have 12 HP heaters, and the problem developed in 8 of the 12 HP heaters after 8 years of service.

To reduce the risk of further tube failures and extend service life, we engineered thin wall tube sleeves that were explosively bonded to the original 304N tube ID surface. The sleeves were 7 feet in length to extend past the desuperheater zone in each heater.
Our thought behind the sleeves was to stiffen the existing tubing and reduce the bending stresses between the segmented baffle plates below the threshold for fatigue crack initiation/propagation.

We qualified the explosive sleeving process with a reputable heat exchanger repair company to assure we had adequate bonding, acceptable heat transfer and did not induce significant plastic deformation to the 304N tube OD because the tubes need to move in the baffle plate holes under various thermal conditions in service.

The cause of the tube leaks appears to be directly related to the heat exchanger designer recommending thinner wall 304N tube material to improve heat transfer characteristics. The tubes were adequately designed for hoop stress conditions but were too thin for use with segmented baffle designs that create excesive unsupported tube spans.

Thus, when ordering new HP feedwater heaters use extreme caution with tube materials like 304N that exhibit higher allowable stress values, in comparison to 304 or 304L, to maintain heat transfer characteristics.

 

[unclesyd]

Here is the website of a company that does a similar lining only they use hydraulics.

http://www.cti-ind.com/

 

[metengr]

Thanks. We did evaluate hydraulic forming of tube sleeves but found that the level of metal to metal contact was not adequate enough for heat transfer purposes.

 

[rmw]

Did you require the OEM to perform vibration calculations on the FWH tubing to check for hydrodynamic whipping in the DSH to verify that the baffle spacing was adequate??

Was the cracking throughout the DSH zone, or was it in a particular area of the DSH zone??

Is the DSH zone single flow, or double flow.

What type of baffling is used?? How are they cut?? vertical cut lines, horizontal, double segmental, etc,???

Are the heaters single stream, or split stream, and if split stream, is one stream isolated on a very often basis?

If you can answer some of these questions it might help with an understanding of what is going on.

rmw

 

[metengr]

RMW;
Thanks for the interest. Prior to your post, we hired a reputable heat exchanger expert who is providing us with some of the answers specific to extraction steam/feedwater flow vibration for the current DSH design. Based on the results from their analysis we may elect to modify the existing heaters or re-tube with a different material or buy replacements.

These heaters were originally purchased under HEI/ASME Section VIII Div 1 specifications. As far as supplemental design analysis for the questions you posed, this was not done. As a typical utility, they went for the lowest cost heaters.

GWG

 

[SMF1964]

I'm intrigued as to how one would get a mechanical fatigue crack to branch? Or are you dealing with a corrosion-fatigue mechanism?

 

[metengr]

SMF1964;
The thought is we have fatigue crack initiation from higher than design bending stresses on the OD surface of the tube. The fatigue crack is circumferential in orientation, and begins as a single crack from the tube OD surface propagating toward the tube ID surface, thus looking like a tight crevice. We know it is mechanical fatigue because only a single crack may develop within a 3 foot tube span with no evidence of surface pitting or secondary cracks. Our water chemistry program follows EPRI limits and is a once thru system. Keep in mind that the OD surface of the tube is exposed to extraction steam and not aqueous conditions.

As with any crevice that forms, given sufficient time in certain wet steam environments, contaminants in the ppb range can concentrate resulting in TGSCC of certain materials. The minor branching happens midwall in the tubing, and increases in severity near the tip. The degree of tight, branched cracks near the tip suggests TGSCC, whereas, I would expect much less branching with CF. Also keep in mind we are dealing with 304ss at temperatures below 750 deg F so oxidation is not really a factor. Either way, we need to reduce the level of bending stresses in the tubing to avoid this problem.

 

[Metalguy]

Agree, CF usually has multiple crack origins on the surface which grow together into one stress-driven crack.

I think you've nailed the problem exactly.

 

[rmw]

If these heaters were made per HEI, then the tube thickness should have been adequate for the service. However, different mfg'r's have different standards for unsupported lengths between baffles in DSH sections. FWH's are among the most competitive pieces of PP equipment out there, so you probably did not sacrifice much by going with the low price.

Much more is known today about hydrodynamic whip or swirl, with the advent of computer flow modelling, but 8 year old heaters should be up to snuff on this consideration.

However, operation can wreck even the highest price, (or highest quality, which ever comes first) heater.

If the heater is designed for base load, as large stations originally were, but operated cycling, then the dry wall margins built into the DSH section can be reduced, and the tubing in the outlet pass can go from wet to dry to wet during the cycling. SCC can show up there, even for heaters with conservative baffle spacing.

That is why I asked if it was single flow or double flow. Double flow would seem to me to be more susceptible to this.

Also, one thing I did not ask, is, if the problem heaters are in split flow streams, or single flow, and if split flow, then has there been a history of operating half the string with the other side isolated. This can put tremendous vibration loads on the DSH tubing, due to reasons I won't go into here, and exceed design vibrations, if an overload factor condition was not spec'd up front.

Are any of these heaters the first heater (extraction point) past the reheat entry to the IP section of the turbine? This extraction point, in some cycles, especially supercritical, while not too high in pressure, tends to be real hot. A turbine that has lost efficiency due to age, and wear, or other factors, would produce hotter than design extraction temperatures at this point. I have seen design extraction temperatures at this point at borderline for sensitization of SS tubes. At this point, the FW is the hottest it is going to be in the heater, and the outer tube wall can be getting right up there.

What about chlorides in your condensate. I couldn't even tell you how many times I have been told, upon asking that question to an upper level plant management person, and one with a plant full of SS tubed FWH's and condenser(s) that they didn't even check for chlorides. One said that they did check for NA, and had noticed that it was running high, but never associated, until that conversation, that the Na they were detecting was associated with Cl (NaCl) that the river they used for CW was draining a part our country that used tons and tons of road salt in the winter, all of which ran off and flowed right by his plant, and through his worn out leaky condenser. Go figure. Anyway, they were having a lot of SCC, and especially in the wet wall/dry wall trouble region. And lots of retubes of 304N stainless.

Hence, the reason for some of my questions. I am still curious about what might have been the root cause here.

Heaters do survive with various grades of SA-688 stainless tubing without suffering your fate.