Did creep cause a valve stem failure?

[Andy22]

We have a high pressure, high temp steam valve with a 2.5" stem, 416 SS material, with an electric actuator. We found the stem stretched like a tensile test specimen just before breaking. I do not think the actuator is large enough to exceed the stem's yield strength, even considering the elevated temp (~500 oC). But the actuator was not working properly and the valve seemed stuck. So what if there was a high load over a long time period, and the stem stretched due to creep? Is this possible? The actuator is capable of a max pulling force of 80,000 lb.

 

[desertfox]

Hi Andy22

Any chance of some pictures of this failure?
I have had an experience of a valve failing, butterfly got jammed in the bore and actuator could not turn it, it was a high temperature valve and the problem was differential thermal expansion.

desertfox

 

[Ron]

Pictures would be helpful as desertfox noted.

For an electric actuator, I would think it would probably trip out before creep would set in, even at high temp.

At 80,000 lbf, the stem stress would be about 16+ksi in tension at normal temp. Any chance your stem material properties were not as specified?

 

[unclesyd]

I'll have to second the post by Ron as we problems of a similar nature on some large valve rebuilt off site where CS was substituted for 410 SS. fortunately this was caught before any operational problems developed.
To stretch a stem the tension load would have to be continuously applied for a period of time. If you loaded it just one time and as the stem stretched you would unload it.
I would check the specifications for the actuator as some of them can apply a very large force on a stem. As posted above something should have tripped the actuator or alerted an operator that the valve was malfunctioning.

 

[Andy22]

This picture shows a manual operator, but an electric actuator was installed when the failure occurred.

 

[Andy22]

This picture shows the damaged/apparently strecthed section of the stem, looking down into the stuffing box/packing area. This valve is currently backseated and locked open but when we get a chance to replace the stem we will test the material.

 

[metengr]

I do not think the actuator is large enough to exceed the stem's yield strength, even considering the elevated temp (~500 oC).

Incorrect. In most cases, if the limit trip is not set properly you can either overload the valve stem in tension or actual torsional overload or fail the actuator housing - I have seen both types of failures at our Power Plants over the years. Stem materials have been 410, 416 and 422 stainless.

So what if there was a high load over a long time period, and the stem stretched due to creep?

Call it what you want, it can be stress rupture if significant load over time is involved versus a oen time event. Testing the material will only provide creep deformation and stress rupture results that you can compare with published values, currently. It sounds to me like you have pinned down the root cause.

 

[metengr]

One final comment, the 410/416 stainless steels show an appreciable drop in ultimate tensile, yield, creep deformation and stress rupture at above 900 deg F service temperatures. This is why most large steam turbine manufacturers limited the use of 410 stainless at 900 deg F. Alloys like 422 can be used safely up to 1025 deg F.

According to my aged USS files and turbine material data, for 410 stainless steel, the published stress to cause rupture in 1,000 hours, at 900 deg F, is 34 Ksi, at 1000 deg F, the stress is 19.4 ksi.

The stress to cause 1% deformation in 10,000 hours at 900 deg F is 24 Ksi and at 1000 deg F is 9 Ksi.

So, if the material was subjected to 16 Ksi tensile stress as only approximated above from a malfunctioning actuator, this could easily result in creep deformation and possible stress rupture.

 

[desertfox]

Hi Andy22

Thanks for the pictures, the one showing the stem in the stuffing gland, am I right in thinking the area of the shaft that seems to have a step in it, is that were it as stretched or I am just looking at a change in section?
I found this pdf on stainless fasteners which gives the creep strength for a grade 416 bolt and supports in general what metengr as stated.(see page 17 in pdf).
I don't understand though, under what circumstances the actuator would generate its full load and maintain it for so long, perhaps you could enlighten me.
It would be useful if you could post any pictures of the failed shaft ie fracture surfaces they could tell us quite a lot.

desertfox

 

[unclesyd]

One possibility possible for the reduction in shaft dia. would be corrosion from Graphite packing. We had several incidences early on when graphite was first used to replace asbestos where we had several corrosion to some valve stems in 650 psig steam service. I also recall one incident where a asbestos graphite packing material corroded a valve stem.

 

[Andy22]

desertfox - yes, where you see the step or change in the stem's cross section - that is where it is stretched. The stem was of constant cross-section along its length when it was new. It has not completely broken yet (no fracture surface).
Very doubtful it is caused by corrosion because the surface is so smooth, and it looks the same all around the circumerfence. Plus, this stem was only in service 7 months and was not leaking.
Thanks for all the info everyone. We are plagued by actuator malfunctions recently.

 

[desertfox]

Hi Andy22

So how could that part of the stem get loaded so heavily?

desertfox

 

[Andy22]

If the actuator malfunctioned and continued to try to open the valve (pull up on the stem) past its fully open limit, there is a larger section of the stem below the stretched portion that hits the backseat. The stretched area is between where the stem is loaded (by the actuator) and the where it is restrained (the backseat).

 

[rmw]

How did the actuator maintain that force on the valve stem?

rmw

 

[desertfox]

Hi Andy22

What does the valve actually do and is it only adjusted infrequently? If the actuator malfunctioned how would it go unnoticed for so long?
If it is a creep failure than as you say it would need to be, more or less fully loaded for quite sometime to cause the damage your seeing,so if you can show that the valve stem can be loaded and go unnoticed its supports your theory. As someone else mentioned the actuator would possibly trip before creep set in, have you looked to see any evidence that the actuator malfunctioned.
My final question the actuator produces a torque presumeably on the thread, so is the actuator pull of 80000lbs a manufacturer's figure or one thats been calculated for maximum torque.

desertfox

 

[rmw]

Hi again Andy22,

Desertfox must have been reading my 'delete' file. I had a lot more in my post prior to his than I actually ended up sending.

Electric actuators in my experience either run or they stop, but they don't load up and maintain a constant force or that is to say a constant driving force. They can drive into a load and with things like spring mechanisms or backdrive preventers, load up so as to for example put a constant load on a valve seat or prevent backdrivng. But when they reach a stall condition, they have to either stall out (open overloads), torque out - mechanisms sense the overload condition and switch the power off of the motor, strip gears or burn up. Not all, but the vast majority of them act after this fashion.

In your case, if creep occurred, then the limit switches that are carefully set at commissioning to make sure that the actuator doesn't drive into the hard stop ends of travel and damage the actuator or driven equipment would then be well out of range, and then one or more of the above conditions would have to occur.

So something doesn't add up about your actuator as to how you arrived at the condition for which you sent the pictures hence my earlier question.

Don't get me wrong. I do recognize that something stretched that shaft. I just question how an electric actuator would do that. Now, once it was done, then an electric actuator would have problems, since the stroke would have changed.

rmw

 

[davefitz]

The use of an austenitic stem within a ferritic cage or stem enclosure introduces the possiblity of creep ratcheting- the differnece in the coeficient of thermal expansion between the austenitic and ferritic coupled with a high temperature range ( room temp at 40 C up to 500 C ) operating) .