-
UREA PLANT CORROSION T. A. Lees
Sherritt Gordon
Fort Saskatchewan,
At our Fort Saskatchewan plant we use the Starnicarbon process.
We attribute our success with this process to the fact that they
have a patented method of air or oxygen injection. In Figure 1 you
will note the stream from the top comes into the first-stage
rectifier. The solution goes through the rectifier, down into the
bottom. and then up into the heater. This is heated to 325~F. It
then goes into the separator, the gas leaving out the top and the
liquid out the bottom. A form of level glass is indicated
there,
The worst corrosion is at the top. on the gas line. The line we
have is 316L and we have had to re-place it about every 9 to 10
months. We are going to change this to 317L very definitely. We did
toy with Hastelloy F, however, we are going to try the 317L. As you
can note the corrosion is marked at this end of the elbow and where
it goes into the rectifier.
First-stage sepa rotor
The first-stage separator is there on the right. It showed bad
corrosion inside from the top down, this was probably due to water
and gases, We had a purge water system going in there for level
indication and it
G3S_
Corrosion
-160 Ib steam
-
CorrOSion 7
Solution
I ~2~:;;;:===::::=J 35 Ib_ condensate ... Solution
Figure 1. Corrosion sitl!S in the first-stage separator and
rectifier.
40
Mines Limited
Alberta, Canada
seemed to trickle down along the side where the corro-sion
occurred. The gas, flowing upward against it, caused the etching as
there was little below the gas inlet nozzle, We thought we had
eliminated this by putting a titanium tube down below the gas inlet
but the corrosion did continue. As a result, in two years time we
had to replace this vessel. The vessel was 316L 3/8-in. thick and
is now 316L l/2,-in. thick.
We have had the new vessel in operation about 7 months. We
examined it in July, 1965, and itls in excellent condition. There
is no corrosion whatsoever.
Rectifier inlet
The next place that is a sore spot is where the gas enters the
rectifier. The nozzle at this point did give us a lot of trouble,
and by the way, while lim talking about the rectifier and your
mentioning 304, we went for two years and noticed a leak on a
thermowell for a good many months and as it started getting bad we
had to take the plant down in a hurry. When we had it down we
examined and found the thermowell was 304 and it had been in there
fOl' two full years. It was re-placed with 316L.
The inlet gas nozzle on this first rectifier that I mentioned
started giving us trouble at the weld. We overlaid this weld in
July, 1963. (The unit was put on-stream October 1, 1962.) Later,
the nozzle corroded underneath the weld, and through, outside to
the re-inforcing plate. We found it leaking out the weepho1e, so we
had to shut the plant down again in January, 1964, and reinforce
this nozzle.
Dish weld corrosion
During our July 1964 regular shutdown, we re-moved the stub
completely and we overlaid the weld with Hastelloy C. This did not
do too badly but if you will notice, there are two dots indicating
corrosion at the bottom dish of the vessel where the weld had
started to corrode. Metal thickness measurements continued to
worsen. In January of this year we had to shut down for other
reasons and upon examining this, we did find that the dish weld was
getting bad, however, we figured it would last until July. We then
decided to replace the bottom half in July, 1965, and did so. We
have now replaced the bottom half of this vessel with 317L 1/2-in.
thick stainless. It was 316L 3/8-in. thick,
Solution short circuit
One other thing that we found on one occasion was a solution
short circuit. We have a second-stage
-
rectifier similar to this first stage. We found a short circuit
due to a blockage necessitating the dumping of the product to the
sewer. It wasn't getting over to our storage at all. This forced us
to shut down. Examina-tion revealed that our second stage heater,
which is comparable to the one shown in Figure I (a little
smaller), was solidly plugged. We got nothing through it. The
reason was that it was full of iron throughout the Raschig rings.
We further found that there was a terrific amount of corrosion in
that first stage rectifier. The Raschig rings had become paper thin
in this first stage. We have taken steps to eliminate this problem
and 1 think we have it licked.
The only other corrosion we have is a slight bit of etching in
the bottom of our reactor. We have always ground this and then
filled with weld. On the next opening we go in again to find the
same thing is occurring. So this time we tried something new. I
don't know whether it is going to work 01' not. We ground where the
etchings were and did not fill them all. About half the etchings
were refilled by welding and we are going to see how these
compare.
Check valves
Another problem occurs with our check valves, which are on three
lines (3,000 lb.) leading into the reactor; carbamate, carbon
dioxide, and ammonia. They get little etchings on them, which I
assume to be corrosion. They offer us a problem because when we
shutdown, we rely on these valves, and if the operators are not
quick enough in shutting the manual isolations, it gives us a
problem by allowing solution to backup and plug the lines. Every
time we get a chance, we lap these check valves. We go to great
lengths on this even though it is very time consuming.
One other thing I might mention here is that we use aluminum
gaskets on top of our reactor and they stand up very well.
We apply Arocoat, which is made by the British American Paint
Co., to our tower walls at the bottom, to any exposed steel, and
all around our pillars in the building, It has pretty well
eliminated corrosion of steel. If you don't get a good bond, of
course, solution
is going to get underneath, but it does a commendable job and we
are pleased with it.
On the top of our wash column we had a 316 gasket, which we
couldn't stop from leaking. We finally resorted to Durabla. We have
used Durabl a on 250 lb. systems and at 200°F with no problems.
Suction valves
In connection with the Teflon suction valves, we use a Hills
McCanna with stainless steel ball and Teflon seat. We previously
had two stainless steel wedge valves that gave us nothing but
problems, and we couldn't shutdown to repair one pump or the other.
We got this Hills McCanna plug-type ball valve and it has been in
service for two years. It has always held and it has never given us
the least bit of trouble.
Another excellent valve that we have found for small jobs like
sample places, is made by Western Associated Engineering Co. W-c
call it the WACO valve, and I heard you mentioning the other day
Nort about the stems and things going. We find the only way to
success-and I'm sure you must know this-is that you must have a
valve where the packing is between the plug and the rising stem
thread. WACO has this.
We have a Dutch State Mine valve with which we are very pleased.
We tried to get this valve for use in our plant expansion, but
delivery was too slow, but we are pleased with the Dutch State Mine
valve. If you have the time to order, is an excellent valve.
Safety valves
In connection with safeties, we have no prob-lems with them. We
have stearn jackets on all our urea safeties, and then on the
discharge of the safety we have live steam continuously directed
into the dis-charge. This keeps the safety hot and urea will not
solidify in the discharge line. We did find one safety valve that
would eventually have given us problems. We noticed that the
pressure was high and the safety did not lift. Subsequent
investigation revealed that the discharge side was full of
carbamate and this is why the safety didn't lift.
DISCUSSION
Pertaining to papers by N. H. Walton and T. A. Lees.
WALTON-SunOlin Chemical: This is certainly in-teresting data.
Looking at my own experience, I visited a Toyo plant at Aruba. I
think there are several being built in this country now but none
actually in service yet. They had as I remember two corrosion
problems there. One was the reactor. It was a lead-lined re-actor.
and every three to six months they had to shut-down and go into the
reactor to make repairs.
When they shutdown they found lead downstream of the reactor in
various places in the system. The lead was in the shape of balls
and they varied in size from the size of shot to the size of
baseballs.
Another problem that they had at that plant was they used
titanium for their let-down line stem, and again they had th,e same
type of results that we had where erosion seems to be a
problem.
BRESS-Foster Wheeler: With reference to the data you presented
on autoclave corrosion, I think it's important that in a historical
reference the intro-duction of oxygen or air permitted the use of
stainless
steel in autoclave. The eadier processes were based on the
complete exclusion of air. Lead, silver, and, in the downstream
parts of the plant, Ampco are success-ful. Lead will react
quantitatively with any air that enters the system, however, if you
exclude ail', it worked perfectly. I believe that zero corrosion
rates were found.
The important thing about lead is its successful application to
the autoclave. We learned it and our clients learned by bitter
experience how to do this. However, once it was done, corrosion
rates went to substantially nothing. The introduction of the
stainless steels, in the exclusion-type processes, where there was
no ail' in the autoclave, were not successful.
Since the conditions are not so severe in the back end of the
plant, this may permit the use of stainless steels. However, you do
not have the same application of oxygen 01' air to the back end of
the plant. In places like valves or valve stems where you cannot
get ail' in, these conditions may give the same type of corrosion
that was experienced in the autoclave when you excluded oxygen and
tried to use stainless steel.
41
-
WALTON: I think the point's well taken here. You have to have
passivated conditions to start with and maintain passivation of
your stainless steel in order to keep corrosion rates down. We have
had a couple of occa-sions where we've gotten a couple of parts per
million of hydrogen sulfide in the carbon dioxide and you make red
prills in a hurry. In our e:l-.-perience, the only way to stop it
is to shutdown the reactor, fill up with water, and bubble a few
cylinders of oxygen up through it to repassivate the liner again,
then you're okay.
BRESS: In those areas where you experienced fairly rapid and
repetitive corrosion such as the vapor line, do you resort or did
you resort to on-stream inspection for metal thinning or did you
rely upon visual inspec-tion during shutdown.
LEES: No, we definitely did not use on-stream in-spection. We
took these pieces apart every time we had a shutdown. We have an
audio gauging method with which we can read through the metal but
we don't rely on it too much when we are on-stream. We don't think
it is that good while on-stream.
Usually, if we have a leak it develops slowly, we know pretty
well whether we are going to be able to last or what we have to do
about it. It is not too hard to shut the plant down to fix it up
with a temporary repair or something of this nature.
WALTON: This is certainly one good thing about urea plants. You
can shut them down and start them up in a big hurry. It's not like
a natural gas reformer where it takes days to shut it down and
start it up again.
REED-Girdler Corp.: I'd like to confil'm a few of the points
that were made earlier. With respect to nickel, I saw an instance
in which a nickel screen was used in a centrifuge for separating
urea crystals from mother liquor in a solution that contained
substantially pure urea in water at a relatively low temperature of
say l40°F. This nickel screen dissolved in approximately one week,
which was quite surprising. It was a thin screen with fine openings
and it encountered high velocity so that you had erosion.
Concerning your mention of crevice corrosion, I think that the
comment that was made here just a moment ago is pertinent to that
point. Protection, or corrosion resistance, of the austenitic
stainless steels depends upon an oxide film on the surface and, in
the operation of the urea plant, it depends upon the con-tinuous
addition of oxygen to the reaction stream. If you have areas in the
plant where the oxygen can be absorbed and there is no fresh oxygen
present, then the stainless in the contact with carbamate and urea
at these high temperatures will corrode. It is a point that needs
to be watched very carefully in the design of the plants to avoid
such areas wherever possible. In some instances it is advisable to
make arrangements to in-ject a small amount of air into such areas.
For example, liquid level control of bottles, etc., can be
protected by the deliberate injection of a small, very small
quantity of air. Perhaps in the Stamicarbon plant injection of a
small stream of ail' into the separator might protect that gas line
by providing additional oxygen. It may be that in the initial
rectifier that the oxygen that's present in the solution coming
from the reactor has flashed out overhead and the liquor stream
going through the heater does not carry enough oxygen to protect
the equipment.
Mr. Walton already mentioned the erosion of the titanium sterns
on the let-down valve. I was going to mention that titanium is very
definitely a comparatively soft metal. It again depends on oxide
film, and under conditions of high velocity where the oxide film is
re-
42
moved, erosion-type corrosion will occur very rapidly. It's easy
to see visually because the color of the ma-terial is completely
differ.ent where the film is absent compared to where it is
present.
Another point I would confirm is that Hastelloy F has proved to
be a satisfactory material as was in-dicated by these tests.
Zirconium has also proved to be satisfactory but again difficult to
obtain. Also, ex-perience in locations in the hot end of the
system, on the high pressure end of the system where 304 was
inadvertently put in instead of 316L, it ordinarily did not stand
up very well.
SPEED-International Nickel: The data Mr. Walton presented
defines both the question and the answer on materials of
construction for urea plants. E:l-.-perience has shown that
corrosion is proportional to the avail-able free oxygen in the
system. The presence of air permits the use of the austenitic
stainless steels which depend on a passive film for maximum
corrosion re-sistance.
Let us review some of the data presented, As an aid to this
review, I have listed the nominal chemical analyses of the alloys
under discussion in Table 1.
TABLE 1. NOMINAL CHEMICAL ANALYSIS OF ALLOYS.
Alloys Fe er Ni Mo Cu Mn S. S. 202 67 18 5 9 S. S. 304 74 18 8
S. s. 316 69 18 10 2.5 S. S. 317 63 19 12 3.5 2 S. s. 329 70 25 4 1
S. s. 309 62 23 13 2 S. S. 310 53 25 20 17-4 PH 75 17 4 4 Carpenter
20 Cb 3 38 21 35 3 3 Incoloy 804 25 30 42 Incoloy 825 32 21 42 3 2
Hastelloy F 17 22 47 6.5 Hastelloy B 5 61 28 Hastelloy C 5 15.5 54
16 Inconel X-750 7 15 73 2.5 Ti Inc one I 600 7 16 76 Nickel 99.5
Monel 400 66 31.5 Illium G 6 22 56 6 6
The successful use of the austenitic stainless steels in this
service requires careful selection of the alloy's iron, chromium,
nickel, and molybdenum con-tent.
For example, Inconel alloy 600, Inconel alloy X-750, Inconel
alloy 700, Nickel, Nimonic alloy 75 showed poor resistance to the
urea-ammonium carbon-ate environment. The reason: too high a nickel
content, too low in chromium, iron, and molybdenum.
Hastelloy alloys Band C also did poorly. The reason: too high a
nickel and molybdenum content and too low in chromium and iron.
The high iron alloys, such as 405 stainless steel and 18% nickel
mar aging steel performed poorly. The reason: insufficient
chromium, nickel, and molybdenum.
The copper-bearing alloys (Monel alloy 400 and Ampco 8), as to
be expected, were severely corroded.
In general, the best resistance to corrosion was demorlstrated
by the austenitic stainless steels ranging from 304 stainless steel
to Hastelloy alloy F.
Note, however, that within this nickel-chromium-iron range are
certain alloys designed for high tem-perature service and were not
necessarily intended for aqueous corrosion service. These alloys
are:
-
309, 310, 330 stainless steels, Incoloy alloy 804, and Incoloy
alloy 90 I.
With the remaining alloys, 304, 316, 317 stain-less steels.
Carpenter 20 Cb 3, Incoloy alloy 825, and Hastelloy alloy F, the
resistance to corrosion increases in the order as listed.
In my opinion, 316 and 316L stainless steels should be
considered the minimum standard alloy for urea plant service. In
areas where chloride ion stress corrosion cracking and
intergranular corrosion lTIay be a problem or where the alloy lacks
general corro-sion resistance, then Carpenter 20 Cb 3, Incoloy
alloy 825, or Ha.stelloy alloy F should be considered.
The reason for SOlTIe of the failures exhibited by Mr. Walton
are clear when we exalTIine the failure on the basis of the
previous discussion.
For example, on the valve stern problem the 316 stainless steel
failed from erosion-corrosion. Type 17-4 PH steel was substituted
in an effort to ob-tain better erosion resistance. Unfortunately,
the heat treatment used for hardening also sensitized the alloy and
failure occurred, not by erosion-corrosion, but by intergranular
attack.
We were also shown a section of 316 stainless steel pipe with a
carbon steel jacket used for stealTI heating or water cooling.
Failure occurred when the 316 stainless steel collapsed in an area;
where the carbon steel was welded circumferentially to the 316.
=========:::::..--~ Carbon steel Jacket A;; "A" J3l6 5.5.
pipe
Figure L Section of 316 stainless steel pipe with carbon steel
jacket lIsed [or steam healing or water cooling.
If you sectioned the pipe at Area A, Figure 1, and examined the
failure, you would probably find one or both of the following:
(I) During welding, the 316 stainless steel could have become
sensitized. Intergranulal' corrosion caused a loss of strength in
the pipe which resulted in a collapsing-type failure.
{2} A circumferential weld of this type results in a high
residual stress. This is aggravated by the difference in thermal
coefficients of e::o..-pansion between the carbon steel jacket and
the 316 stainless steel pipe. Chloride ions in the steam could have
caused the failure by stress corrosion cracking. All of the
ingre-dients for this type of failure were present; i.e., chloride
ions, stress, oxygen, and temperatures above 140 0 F.
WALTON: What is the answ:er on this steam-jacketed piping?
SPEED: Steam tracing will help. That would be the simplest and
the most econolTIical solution. The only other thing you can do is
to go to an alloy that resists stress cracking.
Here again is where your nickel comes into play. The least a!loy
we would recommend for this is Carpenter 20 Cb which has the 34%
nicl~ell then up to 825 which has 42% nickel, and then up to the F
which is 46% nickel.
STOCKBRIDGE-Southern Nitrogen: I can corroborate the collapse
too. We had a similar case. It is our
philosophy that if thin pipe fails, put in a thicker one. We had
a Schedule 10 and replaced it with Schedule 40. So far, we haven't
had a collapse.
WALTON: Have you considered Unitrace?
STOCKBRIDGE: No.
WALTON: We had some aluminum Unitrace in our draining system. It
was a grand and glorious failure in our experience. I wasn't there
the first year that the plant was in operation, and I don't know
what the prob-lems were but I know that when I came in they had
some pieces of it sitting around and they said don't ever ask us to
use it.
STOCKBRIDGE: I want to ask Mr, Walton about the globe valve
problem you had. You said that you did some welding. It wasn't
clear to me whether welding was done inside the valve and you
retained the same body.
WALTON: Yes, inside the valve. You build up a mass of metal
where the seat ring would be screwed in, then machined out a seat
which then is integral with the body of the valve.
STOCKBRIDGE: I've recently seen a new product that Dupont's
coming out with for gaskets. It's a material that is Teflon felt.
They dip this into a solution of Teflon. It looked very good to me.
I haven't tried it yet, but I intend to try it some.
WALTON: Talking about crevice corrosion of ring-type joints, I'm
still uncertain in my own mind what the solution is because these
riug protectors, that we're going to try, will only prevent
turbulence from occurring. They're not a positive pressure seal and
carbamate will intrude behind the ring protector right up to the
ring. So I don't feel as though this is the answer.
Injecting air in between the flanges of the ring-type joints,
while this would certainly do it, is a complicated thing to do and
maintain, therefore, it doesn't seem practical.
MASON-Dow Chemical: You were mentioning the difficulty with
pitting in the tubes on the water side. Have you considered the
possibility of using copper clad on the water side? We have had
quite a lot of corrosion problem that have been almost com-pletel y
eliminated by copper cladding the material on the water side.
WALTON: Copper cladding stainless?
MASON: This was just copper cladding on ordinary steel pipe.
WALTON: Yes, duplex tubes would help here although you get a lot
of different opinions on duplex tubes. It can be a fighting word in
some places. I think I was probably partly responsible for
introducing duplex tubes to the Atlantic Refining Co., and they had
some pretty sad results with collapsed tubes for a number of years.
I think probably the science of duplex tubes has improved quite a
bit today.
CHRISTIAN- United States Steel: I would Hke to add a few
additional comments to Mr. Speed's discussion of the failure of the
steam-jacketed type 316 stainless steel line carrying carbamate
solution. E::
-
steel is not very common. I think it would be desirable to
establish just what type of failure has occurred at the welded
connection. If failure was due to intergranu-lar corrosion, I
believe it's more likely to have originated from the process
{carbamate solution} side, in which case going to a low carbon
grade of type 316 stainless steel might prevent a recurrence of the
prob-lem. Certainly, the use of type 316L stainless steel should be
investigated before going to some of the more exotic constructional
materials that were suggested.
A question I also wanted to ask Mr. Walton con-cerns the cooling
water in his urea plant which pro-duced pitting of type 316L
stainless steel tubes in a high pressure carbamate cooler. Mr.
Walton, are you using a corrosion inhibitor in the cooling
water?
WALTON: Yes.
CHRISTIAN: Do you happen to know which type it is?
WALTON: It's a Nalco treatment. I don't remember right now what
the prescription is.
CHRISTIAN: Is it a chromate base inhibitor and are you
chlorinating your cooling tower water for algae control?
WALTON: Yes. Whenever you get into metallurgy you get into
disagreements. Each of us who either is a metallurgist or takes a
part time approach to it, as I do, thinks that we have evidence to
support our stand. For instance, the remarks that were made about
lead lining in a reactor, there is a reactor that doesn't have
oxygen in it, it is lead lined and it is in trouble. Un-doubtedly
there are other lead-lined reactors without oxygen that are not in
trouble.
ROSENBLOOM-Mobil Chemical: I think it's a very good idea to
inject the oxygen as Mr. Reed suggests, but would this be
infringing on Stamicarbon's patent?
WALTON: Well, we could get into a great big argument about this.
Stamicarbon is not the only one that uses oxygen in their process.
What the patent situation is, I don't know. There are a number of
processes which use oxygen injection into the reaction system. by
some means or another. In the Toyo plant that I saw, the oxygen
injection was downstream of the reactor. I guess the other
processes that I know or have seen have, by some means, oxygen
going into the reactor.
REED: Injection of oxygen at any point downstream of the reactor
should fall outside the claims of the Stamicarbon patent.
CRISTO-Esso Research: We're relatively new to the fertiliz.er
field and we have had our share of difficulties. It follows that we
are very interested in all the discus-sion here. One thing we have
learned is that perhaps the delta ferrite content is important in
316 and 316L and that maintaining a low ferrite content in the
welds of this material is important. I was just wondering whether
this group has had any e}.."pel'ience with this. Do you control the
ferrite content in 316 and 316L? Mr. Walton, In these corrosion
tests, that you ran, do you know what the delta ferrite content was
both in the material and in the welds?
WALTON: No. I would have said that any of the 300 series are
completely austenitic and have no ferrite structure.
On the 329, I just had one piece of information, and that was a
double one which of course is very good but it's impossible to
depend on one reading.
44
Anonymous: You can go back to Mr. Swan's answer too. He had the
440 which is a high chromium 400 series stainless, and the chromium
gives you resistance to this type of environment. This is why his
success was good.
Anonymous: This would be another control that you would have to
specify. We don't deliberately try to get ferrite. This defeats the
whole purpose of the austen-itic structure.
JONES-Canadian Industries, Ltd.: I noted eaJ:lier in your test
data, that, in general, corrosion rates were slightly lower with
the regular carbon grades of stain-less steel than with the low
carbon grades. The higher chromium content of type 309 did not
improve corrosion behaviour, neither did the higher molybdenum
content of type 317 or 317L. It would seem that there is a pattern
here suggesting preferred corrosion of the delta ferrite phase,
since the low carbon grades are more likely to have grain boundary
delta ferrite than the regular carbon stainless steels. The high
chro-mium-nickel ratio of type 309 and the higher molyb-denum
content of type 317 are also likely to result in increa~ed delta
ferrite formation, and this appears to be reflected in their
disappointing corrosion rates.
We did have occasion to require wholly austen-itic type 316L
material, two or three years ago, for a specific strong nitric acid
environment, in which we found this alloy was considerably mOl'e
resistant than type 304L. It was possible to obtain satisfactory
Huey ratings on this special type 316L, and microscopic examination
of welded and sensitized specimens re-vealed very clean grain
boundaries, free from delta ferrite. We bought the material in
heats and paid a premium price for the special control of
composition necessary. It is likely that stainless steel to this
chemistry would do well in urea environments.
A completely austenitic 316L will generally have a lower
chromium-nickel ratio than a 316L containing delta ferrite, i.e., a
17% Cr-14% Ni will be more wholly austenitic than an 18% Cr-lO% Ni
stainless steel. American made 316 is generally in the former
category, whereas European equivalents tend to be in the latter
category.
The practical difference is that the latter category will tend
to corrode more rapidly in urea environments. The delta ferrite
present at the grain boundaries will probably be the anode in a
galvanic corrosion cell, with the austenite grains cathodic. If
there is a large amount of delta ferrite, say, over 10%, then it is
likely that there will be a fairly continuous path through the
grain structure along delta ferrite enabling extensive corrosion to
develop. Similarly, if there is only a small amount of delta
ferrite present, it is unlikely that there is a continuous route
through the structure, and the material will resist corrosion.
Delta ferrite will tend to form in the heat af-fected zone
during welding, and corrosion, is likely to be more rapid at this
location. Since ferrite is magnetic, it is possible that welded
samples of stain-less steels to be used for urea plants could be
evalu-ated this way.
MARCH-Atlas Chemical: We have a Stam.icarbon urea plant, the
same as Mr. Lees. We pay a great deal of attention to ferrite
content in our 316. This is true not only in original materials, as
Mr. Lees will vouch for, but in any welding repairs on the 316. We
ferrite check right as we weld, and if we find a ferrite content
that is higher than we consider permissible we grind it out and do
it again. The big problem that we find in holding the ferrite
content down is cleanliness. We've had to
-
educate our welders 50 that we get ferrite contents that we
think will resist corrosion.
SPEED: We have had the same experience on paper and pulp
digesters where the ferrite content is coated preferentially and on
fatty acid reactors where the 316 is coated preferentially. We're
repairing with Incoloy alloy 825 coated electrode in most cases.
Here it's just a matter of economics. You get a fully austenitic
structure and this coated electrode is about half the price of
Carpenter 20 coated electrode. So you might face a big jump on
alloy costs. Certainly I would think this would offset the expense
of going back and check-
ing and chipping out. This will give you the assurance of a full
austenitic structure.
PRESCOTT-C. F. Braun: The Stamicarbon people do specify a
maximum ferrite content. The problem is to find the supplier who
can meet these requirements. It's very difficult to get somebody to
guarantee the very low ferrite content in a large piece of
stainless steel. The other problem is to measure accurately the
amount of ferrite in stainless steel. If you get ten different
people, or ten different laboratories to make this measurement,
you'll get ten different answers. It's a little bit troublesome to
resolve this point.
45