Dear Professionals,

I work in a power plant in Delta State, Nigeria. We usually have combustion trouble on our gas turbines due to the quality of gas we receive, resulting in frequent cleaning of the fuel nozzles. When the units are shut down to clean the nozzles, we allow the compartment to cool down before the maintenance crew can work on the unit. This cooldown takes about 9 hours. We want to cut down on the cooldown time and are considering forced cooling by installing fans/blowers to achieve this. I need your advice on this practice. How healthy is it for the units? Would you recommend we practise that? Are there other ways we can achieve this reduction in downtime?

 

@Benjye15,

This is not a good idea. Using forced cooling by fans/blowers in the turbine compartment will cause the axial compressor casing to cool faster ("shrink") than the axial compressor rotor--which will likely result in axial compressor rubs, or worse. (And the worse can be very serious mechanical damage, up and including broken compressor blades.)

If you want to reduce the turbine compartment temperature you should consider CRANK-cooling the machine, which means placing the machine in CRANK mode and initiating a START and letting the machine go through a purge and then continuing to let the machine CRANK for a period of time. DO NOT open the turbine compartment doors or run the turbine compartment vent fan(s) or use fans/blowers to try to cool the turbine compartment (which will cause the axial compressor casing to cool faster than the compressor rotor inside the casing--which IS NOT a good thing, especially when the rotor is being spun by the starting means).

How much time? Closely monitor the wheelspace temperatures and when they start to level off (are not decreasing during the CRANKing) initiate a STOP and continue to monitor the wheelspace temperatures. They will actually start to increase slowly after a short period, and when that happens then initiate another CRANK and watch the wheelspace temperatures decrease until they level off again and initiate a STOP.

Repeat this process until the highest wheelspace temperature is around the equivalent of 150 deg F or so. And then you can initiate a STOP. The machine will probably still want to go to COOLDOWN mode (ratchet; slow-roll; turning gear--whatever method is used or whatever it's called at the site) as it's coasting down or reaching zero speed. (I presume the machine is ON COOLDOWN (automatically) while you are cleaning the fuel nozzles, or someone forces it OFF COODOWN.) BUT BE SURE to run the Aux. L.O. Pump (in hand or by forcing the logic signal) to keep the flow of L.O. to the machine's bearings to prevent damage to the bearings because the rotor is still hot. (L.O. flow will help to cool the bearings while the machine is OFF COOLDOWN.) Or, if you're NOT forcing the machine OFF COOLDOWN while cleaning the fuel nozzles, then this isn't necessary (but is recommended IF the highest wheelspace temperature is less than approximately the equivalent of 150 deg F). Once the highest wheelspace temperature is less than 150 deg F the machine can be taken off COOLDOWN while cleaning the fuel nozzles.

Whatever the turbine compartment temperature is during this process is what it will be. The MAIN CONCERN is not to cool the axial compressor casing faster than the internal axial compressor rotor. And using fans/blowers to lower the turbine compartment temperature is NOT RECOMMENDED (this includes opening the turbine compartment doors). NOTE that by CRANKing the machine and forcing cool air through the axial compressor, combustion cans, turbine section and exhaust the axial compressor casing AND the combustion cans (where the fuel nozzles are located) will also cool faster than just cooling the turbine compartment. There is a LOT of heat inside the axial compressor (remember axial compressor discharge temperature can be around approximately 700 deg F or more at Base Load). And it takes time for that mass of metal (the axial compressor rotor is the turbine component with the highest mass and compressing air makes that mass get hot!) to cool if there's not much air flowing through the machine (from the axial compressor inlet to the gas turbine exhaust). And by CRANK-cooling the machine ambient air is flowing INTO and THROUGH the machine, cooling the axial compressor rotor AND the axial compressor casing AND the combustion wrapper and combustors--all of which will HELP reduce the turbine compartment temperature.

It should be clear that CRANK-cooling the machine (sometimes called (forced COOLDOWN) is much better for the machine than just using fans/blowers to cool the turbine compartment.

What isn't know is how long the fuel nozzle removal, disassembly, cleaning, reassembly and re-installation takes. If it takes more than six or eight hours (approximately) then you might be okay with shutting the machine down and taking it off COOLDOWN (as long as you leave the Aux. L.O. Pump running!) and using fans/blowers to cool the turbine compartment more quickly. BUT the unit SHOULD NOT be ON COOLDOWN while doing this! If the axial compressor casing cools ("shrinks") faster than the axial compressor rotor then there could be a rub (where there is contact between the rotating and stationary compressor blades which could result in excessive clearances (which reduces compressor efficiency) and worse--and, again, the worse could be VERY BAD (broken blades).

One of the things that might need to be taken into consideration is the type of starting means the machine has. If it has an induction electric motor starting means you may need to monitor motor temperatures, but not usually. If the machine has a diesel engine starting means that is going to require some careful monitoring because when the machine is being CRANKed using the diesel engine the load on the diesel engine is usually higher than the load on an electric starting motor and some GE-design machines have very marginal diesel engine cooling systems, really only adequate for one or two STARTs (which includes purging and acceleration). So, this is another possible consideration for CRANK-cooling. The good news is--generally diesel engine cooling capacity can be increased (through some additional equipment, of course). (I have seen "cold" water sprayed on the diesel cooling expansion tank help with keeping the diesel engine from overheating--to a point.)

Now, what you're doing (shutting down to clean the fuel nozzles because of dirty fuel) is not solving the root cause of this need to clean the fuel nozzles between maintenance outages. You are losing revenue when the machine is not producing power (during the cooling, disassembly, the fuel nozzle cleaning, and reassembly). Wouldn't the more prudent thing be to get proper filtration/conditioning of the fuel supply to prevent the need to shut down, remove and clean the fuel nozzles, and reassemble the machine? Over time proper filtration and removal of fuel contaminants, precluding the need to shut down to clean the fuel nozzles and losing revenue because of lost production and could even result in increased revenue (over time).

Anyway, hope this helps to understand the potential pitfalls (which could be very serious) of what you are proposing.

 

@Benjye15,

This is not a good idea. Using forced cooling by fans/blowers in the turbine compartment will cause the axial compressor casing to cool faster ("shrink") than the axial compressor rotor--which will likely result in axial compressor rubs, or worse. (And the worse can be very serious mechanical damage, up and including broken compressor blades.)

If you want to reduce the turbine compartment temperature you should consider CRANK-cooling the machine, which means placing the machine in CRANK mode and initiating a START and letting the machine go through a purge and then continuing to let the machine CRANK for a period of time. DO NOT open the turbine compartment doors or run the turbine compartment vent fan(s) or use fans/blowers to try to cool the turbine compartment (which will cause the axial compressor casing to cool faster than the compressor rotor inside the casing--which IS NOT a good thing, especially when the rotor is being spun by the starting means).

How much time? Closely monitor the wheelspace temperatures and when they start to level off (are not decreasing during the CRANKing) initiate a STOP and continue to monitor the wheelspace temperatures. They will actually start to increase slowly after a short period, and when that happens then initiate another CRANK and watch the wheelspace temperatures decrease until they level off again and initiate a STOP.

Repeat this process until the highest wheelspace temperature is around the equivalent of 150 deg F or so. And then you can initiate a STOP. The machine will probably still want to go to COOLDOWN mode (ratchet; slow-roll; turning gear--whatever method is used or whatever it's called at the site) as it's coasting down or reaching zero speed. (I presume the machine is ON COOLDOWN (automatically) while you are cleaning the fuel nozzles, or someone forces it OFF COODOWN.) BUT BE SURE to run the Aux. L.O. Pump (in hand or by forcing the logic signal) to keep the flow of L.O. to the machine's bearings to prevent damage to the bearings because the rotor is still hot. (L.O. flow will help to cool the bearings while the machine is OFF COOLDOWN.) Or, if you're NOT forcing the machine OFF COOLDOWN while cleaning the fuel nozzles, then this isn't necessary (but is recommended IF the highest wheelspace temperature is less than approximately the equivalent of 150 deg F). Once the highest wheelspace temperature is less than 150 deg F the machine can be taken off COOLDOWN while cleaning the fuel nozzles.

Whatever the turbine compartment temperature is during this process is what it will be. The MAIN CONCERN is not to cool the axial compressor casing faster than the internal axial compressor rotor. And using fans/blowers to lower the turbine compartment temperature is NOT RECOMMENDED (this includes opening the turbine compartment doors). NOTE that by CRANKing the machine and forcing cool air through the axial compressor, combustion cans, turbine section and exhaust the axial compressor casing AND the combustion cans (where the fuel nozzles are located) will also cool faster than just cooling the turbine compartment. There is a LOT of heat inside the axial compressor (remember axial compressor discharge temperature can be around approximately 700 deg F or more at Base Load). And it takes time for that mass of metal (the axial compressor rotor is the turbine component with the highest mass and compressing air makes that mass get hot!) to cool if there's not much air flowing through the machine (from the axial compressor inlet to the gas turbine exhaust). And by CRANK-cooling the machine ambient air is flowing INTO and THROUGH the machine, cooling the axial compressor rotor AND the axial compressor casing AND the combustion wrapper and combustors--all of which will HELP reduce the turbine compartment temperature.

It should be clear that CRANK-cooling the machine (sometimes called (forced COOLDOWN) is much better for the machine than just using fans/blowers to cool the turbine compartment.

What isn't know is how long the fuel nozzle removal, disassembly, cleaning, reassembly and re-installation takes. If it takes more than six or eight hours (approximately) then you might be okay with shutting the machine down and taking it off COOLDOWN (as long as you leave the Aux. L.O. Pump running!) and using fans/blowers to cool the turbine compartment more quickly. BUT the unit SHOULD NOT be ON COOLDOWN while doing this! If the axial compressor casing cools ("shrinks") faster than the axial compressor rotor then there could be a rub (where there is contact between the rotating and stationary compressor blades which could result in excessive clearances (which reduces compressor efficiency) and worse--and, again, the worse could be VERY BAD (broken blades).

One of the things that might need to be taken into consideration is the type of starting means the machine has. If it has an induction electric motor starting means you may need to monitor motor temperatures, but not usually. If the machine has a diesel engine starting means that is going to require some careful monitoring because when the machine is being CRANKed using the diesel engine the load on the diesel engine is usually higher than the load on an electric starting motor and some GE-design machines have very marginal diesel engine cooling systems, really only adequate for one or two STARTs (which includes purging and acceleration). So, this is another possible consideration for CRANK-cooling. The good news is--generally diesel engine cooling capacity can be increased (through some additional equipment, of course). (I have seen "cold" water sprayed on the diesel cooling expansion tank help with keeping the diesel engine from overheating--to a point.)

Now, what you're doing (shutting down to clean the fuel nozzles because of dirty fuel) is not solving the root cause of this need to clean the fuel nozzles between maintenance outages. You are losing revenue when the machine is not producing power (during the cooling, disassembly, the fuel nozzle cleaning, and reassembly). Wouldn't the more prudent thing be to get proper filtration/conditioning of the fuel supply to prevent the need to shut down, remove and clean the fuel nozzles, and reassemble the machine? Over time proper filtration and removal of fuel contaminants, precluding the need to shut down to clean the fuel nozzles and losing revenue because of lost production and could even result in increased revenue (over time).

Anyway, hope this helps to understand the potential pitfalls (which could be very serious) of what you are proposing.

Thank you very much @WTF? for the detailed explanation and recommendation. We are currently procuring filters and other gas defence equipment, but with a long lead time. A short-term solution is what we need. Are there other ways we can reduce downtime for the Nozzle cleaning exercise?

 

@Benjye15,

I've given this a lot of thought, and to my mind CRANK-cooling the machine is the best for the machine.

Actually, using the method I outlined of starting and stopping CRANKing based on wheelspace temperatures is the fastest method of getting the machine off COOLDOWN. Ambient air being drawn into and forced through the machine as it's CRANKing cools the entire machine faster (internally as well as the casings and combustors).

Again, the thing you don't want to do is cool the axial compressor casing faster than the axial compressor rotor--ESPECIALLY if the machine is ON COOLDOWN. That's just a recipe for disaster because the likely outcome of using fans/blowers the cool the turbine compartment (where the axial compressor and turbine are located) while the machine is ON COOLDOWN is that a rub WILL develop and it could actually prevent the cooldown sequence--causing the machine rotor to be "stalled" (unable to turn because of contact between rotating and stationary blades). Once that happens, Management and Supervision and probably the Mechanical Department will be frantically urging to use external methods to turn the rotor--which has caused catastrophic damage to more than one machine during my career. (It cost one site to loose 6 months of revenue while repairing the axial compressor rotor and casing including replacing many rows of rotating and stationary axial compressor blades. Because the site produced steam for a nearby food-processing plant mobile boilers had to be brought in to provide steam to the plant (they had some standby auxiliary boilers but they didn't produce enough steam--at an additional cost to someone, I'm sure).

Anyway, the only thing that might be helpful is that as the wheelspace temperatures drop (signifying the axial compressor rotor is cooling, as well as the axial compressor casing and the combustors) you could start the turbine compartment vent fan(s) and maybe open one of the doors near the compressor inlet to help with cooling the axial compressor casing and combustors a little faster. BUT, I wouldn't do this during initial CRANKing for the reasons already described. That MAY help with cooling the compartment a little faster by possible increasing the air flow over the axial compressor casing once it's started cooling down.

(I'm presuming the machine has turbine compartment vent fan(s) located on the turbine compartment roof, and not the tube-axial fan in the bottom of the machine. If the machine has the tube-axial fan in the bottom of the turbine compartment it draws air from the Accessory Compartment, which is usually pretty warm (warmer than ambient air anyway) and might not cause an excessive cooling of the axial compressor casing. But I wouldn't open any turbine compartment door until the machine internals are well into lower temperatures by the flow of ambient air through the compressor and turbine and exhaust.)

I've had to make several presumptions because we don't know what Frame size machine is at your installation and we don't know how old/young it is and we don't know about the turbine auxiliaries (cooling fan(s)--either one or two roof-mounted fans, or a tube-axial fan in the bottom of the turbine compartment that draws in from the Accessory Compartment. It would really help if you could provide some basic information in the original post so that it's not necessary to assume anything and to try to cover all the possible configurations for turbine compartment cooling, at least.

Anyway, that's all I got.

It's a world-wide problem--dirty gas being delivered to natural gas-burning machines. Some places it's really bad; others a simple knock-out drum works fine as long as it's blown down periodically and cleaned during planned maintenance outages. I've seen sand, silica, scarf (metal drill turnings from in-site piping repairs and modifications--some of which get through the y-strainer at the entrance to the gas control valves), broken drill bits, broken springs, diesel fuel, gasoline and compressor lubricating/seal oil (the natural gas compressor lubricating/seal oil), filter pieces, rags and cigarette butts (which had almost completely blocked the y-strainer forcing a shutdown to remover the internal screen basket to get the butts out). And that's in first-world countries as well as lesser developed nations, also. At one site broken springs (from where we couldn't really tell) prevented the incoming gas fuel pressure regulator from working properly and the machine had tripped on high-high exhaust temperature. EVERYONE thought the problem was the turbine control system because the machine couldn't be started (it would trip on loss of flame OR high-high exhaust temperature on START attempts). And it was the incoming gas fuel pressure regulator plug that was prevented from closing and limiting/controlling gas fuel pressure to the site. And, of course, there are the problems with high-high exhaust temperature spreads caused by fuel contaminants (including natural gas liquids which condensed as they passed through the control valves and the fuel nozzle orifices).