proving reverse power protection relay is working through real reverse power 17

[jimmy2times]

We have a couple of diesel generators and gas turbine on a ship. The class society guy wants us to prove the integrity of reverse power relays as part of his intermediate (2.5yr) survey. He suggested other operators choose to actually induce a real reverse power situation by taking load off one set and driving the set frequency up on the other so as to motor the unloaded generator. What is general view out there on this? I would have thought less risk of damage to just get a contactor in and make protection tests on the relay without the generator running.

 

[waross]

I disagree. The circuit may be wired perfectly but the settings are wrong for the application.
The real world test proves the settings as well.
If you hesitate to do a real world test, then consider the possible consequences of an actual real world reverse power when the reverse power protection fails to operate due to improper settings.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[marks1080]

waross real world test do not necessarily prove settings are correct. It may prove you're 'in the ball park' but it isn't going to give you good insight into the exact pick up levels of the quantities being measured.

If the goal is simply proving the relay is working and healthy the most appropriate way forward is testing the relay with an injection test set.

If the goal is proving the system as a whole works than you want to look at doing real world tests.

Typically a maintenance cycle on electrical protections begins by testing all the individual components (ie: relays) and ends with an overall test of the system based on a real world event.

All of my experience is with utilities so if there are other industry practices for oil&gas I would be unaware.

 

[catserveng]

I think Maritime regulatory agencies are pretty much always concerned with testings that demonstrates system response and resilience. I just had another conversation about a similar issue on a newer diesel electric research vessel. The diesel inspector is concerned that a single engine failure on a multi-unit bus does not cause a loss of propulsion or house loads. When you talk to most of these guys the fact a single unit fails isn't their main concern, it is how that failure affects the vessel systems. A single unit on a bus fails, plant goes dark but next available unit comes on, most times automatically pretty quickly with minimal operator interaction. If two units are on the bus, one fails but it doesn't clear quickly enough and causes the second unit to fault, now there is a bigger problem that may take more time to resolve.

Some of these inspectors can be very hard to please, and your arguments need to be pretty darn good to get them to accept something other than what they think they need to see. But most of these inspectors were also working engineers with lots of experience who in just about every case I dealt were passionate about getting it right.

Last year I dealt with 9 separate instances of units failing to trip on reverse power after an engine control or mechanical failure. In all cases the relay performed as expected, asserting the trip output when tested using a test set. In all cases the setpoint used was too high for the actual parasitic power it took to drive the unit when paralleled to the bus. In 7 of those cases we were able to go back to the engine manufacturer and get a estimated parasitic power number, in two cases we could not and in both cases, a "real world" test was done to see what the reverse power level really was and develop an appropriate trip setting. In two of the cases where we got a number from the supplier, the end user insisted on a real test due to the amount of damage sustained by the units for not tripping (the units had been driven for several hours before being discovered). Also in 6 of these cases the electronic engine controls or monitoring system were not properly integrated with the breaker controls, so instead of getting a proper signal telling the switchgear the engine died, the switchgear controls relied on the reverse power setting to open the breaker.

Every industry segment has its own quirks, utilities have different concerns than ships, even in maritime what the ship does has a big impact on how the regulatory folks deal with them and to what extent they will test things out. At from what I've seen

MikeL.

 

[waross]

Last year I dealt with 9 separate instances of units failing to trip on reverse power after an engine control or mechanical failure. In all cases the relay performed as expected, asserting the trip output when tested using a test set.

Thanks for that insight Mike.
The difference between making sure before or making excuses after.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[marks1080]

Real world tests are good and should be done.

More important, imho, is having a documented maintenance process that covers everything and isn't reactionary.

Should the specific relays be tested individually (ie: bench tested)? Absolutely. Should real world tests be done as well? Absolutely.

Should we understand the purpose, impact and difference between these things and why they are both needed to prove overall system health? Absolutely.

So... if the question is whether or not to do secondary injection tests on individual relays, or to just to overall high level system tests ('real world tests') the answer is yes. To both.

This person seems to be required to provide proof a relay works. The appropriate thing to do is to test the relay. I'm assuming that system level tests are required as well for other purposes. To simply fire off a real world test for any maintenance scenario is not a good way to test protections.

 

[jimmy2times]

I had the feeling surveyor would be more content in seeing gauges operate and the set trip rather than a record of the unit being injection tested. I guess putting myself in their shoes i would want the same mainly as we could get it bench tested today if we wanted and just show him the test record and the credientials of the contractor undertaking the test and he would have to take this as evidence of the test. So seeing it work there on the spot or having some paperwork pushed under your nose on the next visit i could see why he was asking for this. Will look at best way to demonstrate it to him next time and establish that settings are correct in meantime, thanks for your assistance yet again.

Off subject, but a document you may be interested in mike ......

http://www.globalmaritime.com/library/media/1486108100.0096.pdf

 

[catserveng]

Jimmy,

Great article, DP has always been a really interesting topic, where it started being pretty loose to holding really tight, especially for directional drilling systems. A few years ago ABB also had a very good paper about open vs closed bus ties on larger marine power systems, fast operating sectionalizing schemes seemed to play a big role in those ideas. I don't do much of that work anymore and I think it has changed a lot in the last few years. I know we used to keep a lot of units on line at low loads to support open bus tie schemes, but the newer engines or in some areas having to comply with emissions regulations started to make that less attractive. Some operators mixed schemes based on what the load profiles actually were, less critical loads staying on with fewer generators trying to load them up, or adding ballast load in some cases.

When I started doing a lot of offshore work, the inspectors were brutal, especially to us "smart ass american whiz kids" as some of the old DNV guys used to call us. Was able to get to a point where we could actually talk about their concerns, what the end result of the inspections needed to be, what steps in testing did they want, and what areas did they feel they got mislead before. I think what has changed is that we had more time to actually talk things thru, a lot of this was newer and we all were learning together. It seems like no one has the time or desire anymore to actually work it out together, just get it done as fast and cheap as you can and get back to work seems to be the goal.

Good luck in your venture, thanks for sharing the article, was a great read.

MikeL.

 

[jimmy2times]

No problem thought you might take interest in it following on from your discussion on fault tolerance

 

[waross]

Why the reluctance to do a real world test? The fastest and easiest test is to just dial down the speed on the set under test.
When you consider Scotty's description of the damage that may be done to turbines by reverse power and Mike's experiences with failed reverse power schemes, give the inspector what he wants.
You will be done in less time than it will take to even plan a bench test.


Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[jimmy2times]

I am in agreement and will look to do the reverse power as live test for the surveyor and gain confidence in the system rather than isolated bench test, just want to first establish that the relay is set correctly first. Agree we can then demonstrate this quickly to surveyor each time (his preference yearly demonstration but his minimum requirement as intermediate surveyor which is 2.5yrs) so i agree this will be much better test and less effort and planning required than testing relay in isolation each time.

 

[ScottyUK]

I would want to be as sure as possible that the relay settings are correct prior to a live test by enquiring what the expected absorbed power is for the prime mover when driven by the load, but once you're at that point a live test isn't unreasonable. The prime movers aren't going to fail immediately that the power swings from generating to motoring, it takes a little while before damage occurs. Make sure that you are ready for a manual intervention in the event that the 32 relay doesn't operate as expected, and don't linger too long in making that decision to trip the set manually.

If your control system incorporates a Historian, or if you set up a Dranetz or similar power analyser, then you'll be able to see what the real reverse power load is during the test, data which will help you immensely if your relay doesn't operate because the prime mover doesn't require quite as much motoring power as the manufacturer thinks. I imagine that a diesel engine being driven at rated speed consumes a fair bit of power so it should be reasonably easy to establish usable settings.

 

[waross]

I imagine that a diesel engine being driven at rated speed consumes a fair bit of power so it should be reasonably easy to establish usable settings.

That depends a lot on the individual engine model.
Years ago there was an engine that was commonly used in transport trucks that was noted for almost zero hold-back on down grades. This engine had so little hold-back that the time to lose enough RPMs to synchronize an up-shift was many times the time of competing engines.
A setting that works for 98% of diesel engines may not work for all engines.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[crshears]

Kind of where I was going in my posts, Bill; for sure total losses in Diesel engines on line & at speed but at zero fuel rack can vary significantly between engine types.

Although I take your earlier point about Diesel engines not liking to run at low or zero load for extended periods [as this almost certainly will cause fouling], reverse power tripping commonly occurs with no intentional time delay, rendering this observation, with respect, not applicable to this instance.

It is my understanding that once the equipment need to perform the type of data capture ScottyUK is suggesting [whether it be that type of equipment or even the more primitive method employing temporary raised-zero reverse-reading-capable meters[ is in place and reverse power protection has been blocked, the process of forcing fuel rack to zero, recording readings, and releasing fuel rack for normal operation could all be accomplished within at most thirty seconds; although one might exist, I do not know of any Diesel engine that will foul in that short a period.

And unless I'm way off base, if the reverse power prot on an in-service gas turbine is blocked beforehand, the sequence of forcing fuel flow to zero, capturing readings and manually tripping the unit out of service to normal shutdown should be capable of completion within a similar time frame, with no resulting damage [anybody who knows otherwise, please say so; I'm still learning].

CR

"As iron sharpens iron, so one person sharpens another." [Proverbs 27:17, NIV]

 

[ScottyUK]

Hi crshears,

On a single-shaft machine the power consumed by the compressor at sync speed is so large that workable relay settings are almost guaranteed - you're looking at a load of about 30% of the power turbine's gross output: put another way, the compressor consumes the equivalent of half the set's electrical output. That's a hell of a big target to miss, even with calculated settings.

You could probably do as you suggest with a GT, but I haven't seen it done that way. You'd have to defeat a fair bit of the control logic in order to run that test.

 

[catserveng]

For years working around the engines (gas and diesel) and generators I mostly deal with in the .5 to 5 MW range, the default or rule of thumb type setting for reverse power was 10% of rated for 10 seconds. You saw it not only in the protective relays but in multifunction load controllers like the Woodward DSLC and Basler DGC-2020.

Now look at all the improvements in horsepower to weight/size ratios in modern gas and diesel IC engines, a 500kW genset was a big chunk of iron, and not just the engine but the tail end as well. And internal improvements like smaller bearing/crankpin surface area per BHP than older engines, improved piston ring to liner interface, improved cam to lifter profiles, all this stuff made the engines frictional horsepower go way down. But did that really get passed on down the trail to the folks doing the electrical protection? Do the current modeling softwares like SPM and ETap take that into account? I have an older protection handbook that states "Maximum motoring power of a diesel engine is 25% of rating", that was probably a good guideline for a large slow speed engine many years ago, not sure it would be of any use today.

I still see a lot of units in the field with 10/10 reverse power protection, on a high speed highly turbocharged engine and a smaller, lighter generator end, I will bet money that engine has nowhere near a 10% of rated parasitic horsepower demand. But they still showup, and almost every time I bring it up it, it becomes an argument. Of the 9 units I mentioned last year, I was involved on three of them early in their service life. In all three cases I expressed my concern, in all three cases the settings were not changed. On top of the fact the engines have reduced frictional losses, modern engines also have more extensive protection system with very large numbers of active shutdown protections. Take an "old" style mechanical diesel engine with a standard protection package, it likely had low oil pressure, high water temperature, overspeed and emergency stop shutdowns. It was also in most cases an Energize to Shutdown protection system, so a loss of control power would allow the engine to continue to run. Take an engine driving a similar size generator today with modern electronic engine controls operating under some sort of emissions compliance. I would bet at a MINIMUM there are at least 20 active shutdown protections, and if control voltage drops while the engine is running it will operate erratically or shutdown altogether. I'm not sure if this is an issue on the smaller end of the size range or because it driven by and IC engine. Sounds like the gas and steam turbine folks are more diligent in their protection schemes.

Nice thing about a forum like this is the differing perspectives everyone has based on what they work around, not everyone may have all the "right" answers but it sure brings in a lot of great info.

Hope that helps, MikeL.

 

[waross]

The issue with diesel engines and light or reverse loads is piston ring seating. Many engines are designed so that the pressure of combustion acts to increase the pressure of the rings on the cylinder walls. With lightly loaded sets, there may be insufficient pressure to properly seal the rings and the result is oil pumping or oil loss at low loads. This can also lead to cylinder wall glazing, which may result in the oil loss continuing when the load is increased.
Also, most new sets have the rings seated well, but the odd set will not have the rings set and will pump oil under light loading.
I installed a set for a meat processing plant. They had one large machine called the cutter that ran for a couple of hours each day. The genset was flat out, blowing black smoke and leaning on the mounts when the cutter started. Then it dropped down to about 20% to 25% of rated load. When the cutter was not in use the load may be as low as 10% of rated load.
When the dealer became aware of our load profile he demanded 5 days on a load bank to ensure that the rings were properly seated.
(With a 275 kW set I sure boiled a lot of water that week.)
The damage from reverse power on a diesel may be long term and is often reversible by simply loading the set.
Consider the Duromax diesel engine. GMC announced back in 2007 that they had just built the 1,000,000th Duramax engine.
By now they must be close to 3,000,000 engines. Every time the driver takes his foot off the gas, the engine is in reverse power mode. In the mountains the engine may be in reverse power mode for miles at a time.
No-one has problems with a Duramax engine due to coasting down hills.
I don't see any problem with the majority of diesel engines in reverse power mode for a few seconds or even a few minutes until the protection trips.
There is the rare exception. That is some old worn out standby sets. These are in such bad shape that they wouldn't last more than a few months as prime power sets but the owners won't replace them until they fail completely.
The rings and cylinders are worn so badly that the pressure of combustion is the only thing sealing the rings.
Under reverse power they will pump oil so badly that there is an actual danger of throwing globs of burning lube oil out of the exhaust stack. Still the reverse power does not damage these engine any further until either they run dry of oil (seen that) or set fire to the generator shack. (haven't seen that yet but I have heard stories)
A diesel engine is not subject to the kind of immediate, expensive damage that a gas turbine is under reverse power and there is no danger to the engine by a real world test.
A suggestion. If you have instrumentation installed that will measure reverse power, disable the reverse power protection and see what it takes to drive the set. Then enable the reverse power protection and check that the set point is well below the maximum reverse power. If the set point is too close to the actual maximum reverse power the action may not be reliable.

Bill
--------------------
"Why not the best?"
Jimmy Carter