Several MarkV sites have encountered fires related to a shorted solenoid valve output (generally powered from QD1 DTBD terminal.) These outputs are not individually fused so a short at the solenoid can and has caused the terminal board to overheat and burst into flames. It is never good to let the smoke and fire out of a wire or circuit board. Several sites have installed fuses on these outputs, which is a good idea. The Mark6 has these fused. We are interested to know if anyone else has done this.

As I look at doing this project, I find I am not sure what size fuse to use. Other sites just install a 3amp fuse. I have listed all the solenoids and have 'most' of the amp ratings. My question is; What size fuse should we install on a solenoid valve rated at 125VDC, 35.1watts, and 0.28 amps? Is there a rule of thumb here? Just slapping in a 3 amp fuse is not logical to me.

So far I find that the Asco web site states that a DC solenoid has no inrush, so I should just fuse at the rating, but for the above example, I find it difficult to assume a fraction amp instantaneous fuse will be acceptable.

 

Most of these incidents can be traced to improper wiring practices. It's difficult to describe without graphics or pictures, but for most of the heavy duty gas turbine Mark V panels GE packaged they did not typically use two wires for every solenoid output (a "hot" wire and a "common", (wire no. (108)). Instead, they usually use a single common wire from the Speedtronic to a junction box on the unit that might have two or more solenoids connected to that junction box, and then jumper that single common (108) wire to all the solenoids connected to that junction box. Unfortunately, it's "common" to find solenoid outputs in the Mark V with only a single wire (on the NO terminal) and no wire on the SOL terminal (refer to the Signal Flow Diagrams in Appendix D of the Mark V Application Manual, GEH-6195, for more information about output terminations).

What <b>SHOULD</b> have been done, either in the factory or in the field, is that the SOL terminals of outputs being used as 125 VDC solenoid outputs should be jumpered together on the DTBCs and DTBDs and to the limited number of (108) wires that go out to the various junction boxes. That way, all the Mark V outputs configured as 125 VDC solenoid outputs will "share" the current load, instead of only one SOL output providing the current load for multiple solenoids.

GE uses, across the board, 3.2A "slow-blow" fuses in the Mark VI and Mark VIe, regardless of solenoid rating. These fuses will "ride through" a high inrush, but in the event of a fault (short) they will protect the circuits on the terminal boards, which is all the fuses are really there for anyways--to protect against a fault.

If all you're trying to do is protect the Mark V terminal boards (and the ribbon cables connecting them to the TCRA cards, and the TCRA cards), then a similar fuse will do the trick. But <b>ONLY</b> if the SOL outputs are jumpered together at the Mark V <b>and</b> to the limited number of (108) wires that are run out to the field! Or, if you run individual (108) wires for every solenoid output.

If you're having this problem with solenoid outputs that actually have two wires per output, then there's something causing high current flows that shouldn't be. Either the in-rush is too high or the solenoids are not properly rated. You even stated that the problems were caused by shorts, which, under normal conditions, should not be occurring. One would have to ask, why are so many shorts occurring? Is it poor wiring practices? Faulty solenoid coils? Moisture/water getting into junction boxes or coil enclosures? Excessive heat causing insulation to melt or degrade? A lot of times problems like this are not just the result of one issue, but get attributed to one issue. I'll agree, the practice of using a limited number of "common" wires for solenoid outputs isn't great, but it's been done for decades on GE-packaged heavy duty gas turbine applications, and without a lot of problems unless there are field/maintenance issues. Truly analyze each of the shorts/failures to determine if there's something which can be done to prevent future occurrences, by fixing that condensate leak that drips directly on the solenoid coil cover that wasn't fully screwed down; or completely closing the junction box covers; or fixing the conduit(s) that were run into or out of the tops of junction boxes that are exposed to rain/snow and not properly sealed, allowing water to enter the junction box; or poor conduit fittings; or conduit and/or fittings that allow water to enter the junction boxes that have broken because people use the conduit for ladder rungs (but that never happens, right? Pipefitters would never do that, would they?).

Most of the DC solenoids that GE uses do not draw a very high in-rush current or even use much holding current. The exception to this are the Laurence Compressor Bleed Valve solenoids and the Laurence Liquid Fuel Forwarding Stop Valve solenoids. These large DC solenoids do have high pull-in currents, and are supposed to have internal means for reducing the holding current. The 3.2A slow-blow fuses work perfectly for these solenoids. because they allow a high in-rush for a short period of time (fraction of a second) and still protect against a short, or even a problem where the internal current-reducing method fails.

Some of the older Laurence solenoid coils were also polarity sensitive, though they were very poorly identified. The positive lead, which was black--the same color as the negative lead--had a small piece of red heat-shrink tubing somewhere along it's length. Quite often it slipped off or was cut off when the leads were trimmed to length.

So, examine the solenoid output wiring on the Mark V. Make sure that more than one or two or three outputs are not providing all the "common" output current through those limited number of (108) wires that were run out to the field. Every solenoid output that's configured as a solenoid output should have a wire on the SOL terminal, even if it's just a jumper to another SOL terminal that eventually jumpers to a field wire running out to some junction box on the unit. (Again, I'm presuming you're referring to GE-packaged heavy duty gas turbines with typical GE wiring practices.)

The lack of proper jumpering for the (108) terminals/outputs was the cause of the most of the problems you are describing. It's not a shortcoming of the Mark V, it's poor wiring practices when applying the Mark V.

If you find that only a limited number of SOL outputs have field wires, and that they are not jumpered to the other solenoid output SOL terminals and eventually to one or more field (108) wires, then add the jumpers and the problem you have described will basically be solved. (I'm not talking about jumpering EVERY SOL terminal to every other SOL terminal; only the SOL terminals of outputs configured to be 125 VDC solenoid outputs!)

Hope this helps!

 

Thanks for the detailed response, it is helpful.

There were only a few unrelated sites around the world that had the failure but that raised a flag for the rest of us.

I did review the installations here and found that all the solenoids use 2 wires all the way to the Mark5 panel, no jumpers. I also inventoried all the solenoids and charted the amp rating of each. Most are an Asco valve that draw less than .5 amps / 125VDC (with a couple of exceptions.)

Appendix D shows that there is a common power feed to the terminal board and it has a 15 amp fuse. All solenoid outputs are protected by the one 15 amp fuse (if the jumper selection is set for internally supplied.) The worst case of a solenoid or cable that shorts to ground should only draw up to 15 amps then blow the power distribution fuse. I don't know if 15amps would be enough to blow the trace and parts off of a termination card and set fire to cables and cards nearby (doesn’t seem likely to me, and I have blown things up before.)

A post review of a fire at one site states:..'over current event which caused failure of the DTBD card to the extent the card and backing plate were melted. The calculated current is 75 amps'...' The LRSV coil shorted would have caused a high current and direct short on the DTBD card near terminal 25 and 26.’ I don’t know if that site found jumpers on the 108.

One issue I have is how they could have seen 75amps (or anything greater than 15amps?)

One exception to my review is that a couple of outputs do not use internally powered source, they are just contact outputs with power fed from some external source. In that case if there was a short and that external source wasn’t fused properly, then we could find a high current potential running through the terminal card and relay. That may be enough to flame.

With all this being said, I have come to some conclusion:

1. It looks like I will need to define the power source on those few outputs to ensure proper fusing.

2. Since GE is now fusing the outputs on the Mark VI it does seem like a good idea to just add the fuses. It may be overkill but is cheap insurance.

3. It looks like a 3.2 slow blow is what is recommended, now I must decide what type of fuse holder and whether we will use blown fuse indication.

 

Mike,

I am not certain, but we have recently received a report of a fire in a MKV which has very similar details to your references.

We have also been asked to review the report and make an engineering assessment.

During our site's life, we have experienced a couple of failures on termination cards, primarily with the DTBC card. However, each of these occurrences was after performing maintenance works, so there has been an assumption that these have been caused by improper installation.

At our site, we do have jumpered common wires and hence, our original installers, advised us after one of the issues, to load share across QD1 and QD2, however, GE have subsequently responded that this was not the way forward and we have since removed the wiring.

We haven't had any any further failures, but the report we have received has got people talking. Certainly given that the equipment is not getting any younger!

Do you know if GE are looking to release any kind of TIL regarding this? Obviously, as CSA has stated, the root of the problem is the improper wiring, however, with our current position re-wiring would not be a feasible option. We have previously asked GE about fused terminals, but they did not feel that this was required.

Regards
Martin

 

In my personal opinion, it doesn't take 15 A to burn the traces on a DTBC or DTBD, or the ribbon cables connecting them to the TCRAs. And, once there is damage on the DTBC or DTBD, it's not very far, physically, to a trace with the opposite polarity which can cause the arcing and "fires" that have been reported.

I don't know what good a TIL would do unless GE were "offering" a new, improved DTBC or DTBD, that would require one to purchase them. (Aren't those things great sometimes; they warn of failures and get the owners to pay for the remedy(s).)

I was on a site once where the generator breaker trip circuit was not properly protected with a contact to open the circuit when the breaker opened. (This circuit was externally powered, not a typical internally-powered solenoid output.) When the circuit was energized the current drawn by the trip coil was about 10A and it caused a failure of the DTBx card (I don't remember which one). The resulting smell from the burnt printed circuit board was enough to make a person violently ill, and could be smelled in the PEECC for weeks afterward. I'll never be able to get that smell out of my memory.

The 10A current flowed for about 1-1/2 to 2 hours as best we could tell, before the failure occurred. So, it was likely a combination of high current and the resultant heat over time that led to the failure.

Another cause of failed DTBCs and DTBDs is when someone inadvertently puts the 125 VDC jumpers in place on the DTBx on an AC circuit, such as the control circuit of a motor starter. There is usually a 125 VDC battery ground present for some time prior to the point at which the motor starter control circuit is energized, and then it goes away when the DTBx fails, and usually takes out one or more of the MOVs out on the TCPD card in the <PD> core with a loud POP! One can almost understand how this happens during commissioning, but when someone just decides to do this during a maintenance outage (install jumpers which weren't there for years, just because some drawing says they are supposed to be there, and without checking to see what the output does), well, that's just unbelievable. (And it's happened more than once in my experience, so it's probably happened many more times than those.)

 

Martin,

Another user on the 7FA users group commented about engaging GE with this issue. Here is their comment:

"After conducting a very small query, there are at least 3 other known incidents identical to this one. Please share feedback if you have had similar experience. We hope that sharing this information may prevent a similar incident at other sites. This impacts safety as well as production.

2010 PAC case - opened by our sister site "Resolution to PAC case 20100225-0056 has been issued. Recommendation:
The only way to prevent a similar issue from occurring would be to install individual output fuses on each of the devices. For example a
separate set of fusing for each of the solenoid coils."

2005 PAC case - site unknown - PAC case 20051107-003 recommended installing in-line fuses for each solenoid circuit. A fix should be implemented as soon as possible. A solenoid short can still occur and cause damage.

This issue should be a TIL and should have been released some time ago!!!"

It appears that GE has recommended adding fuses but will not commit to a TIL.

You need to fuse both wires which is difficult if they are jumped together.

It sounds like the main problem is a result of these jumpers and I would also recommend that this should be changed to run individual pairs to each solenoid.

We have over 40 solenoids per unit which would mean 80 fuses, that seems extensive. We do not have the jumped wires to deal with so I have some hesitation to do any of it...

Again maybe it would be cheap insurance.

I can't find a slo-blow fuse rated at 3.2amps 125VDC that fits a fuse holder with blown fuse indication.

 

Hi there,

Instead of trying to increase the fuse ratings, you might want to look in a different direction to make this system safe so another possible solution might be to install only small fusses in the output cards and take these outputs to low amp interposing relays at a different location. We normally install all these interposing relays in one of the RTU’s in a safe, easy accessible location.
Route the supplies to these high amp solenoids then from a separate protected high capacity electrical source through the contacts of these interposing relays instead of directly from the output card. This will protect your cards and control system from any possibility of fires due to short as well as high amp meltdowns.

Two negative points I might add is it will be a big job to do a modification like this since all solenoid cables and wiring have to be rerouted and redone as well as the question of getting permission to do a modification like this from GE. If you can get past these this might be a possible solution to protect and make the system safe.

 

A good point about someone adding a jumper, for whatever reason.

During my review of this situation, I also made a mapping of all these jumpers, just in case.

 

Sam, thanks for your reply. Having a higher current go through an external relay is a good idea.