Cooling water system

J

Thread Starter

jaafar

Dear Sir

What are causes if the air accumulates in cooling water line specific in leg cooler and flame detector cooling?? Also because the air in the cooling water the heat increase and flame lost.

We are check the purge air cooler but there is no leakage

regard
 
Air will collect in high spots of piping if not periodically vented. The cooling water jackets of most turbine support legs have 'vents' which can be opened to relieve most of the air which will accumulate in the jackets.

As for flame detectors, well, most of the installations were not provided with similar vents, the thought being that one could loosen the compression (Swagelok) fittings and bleed the air out when there was pressure in the cooling water system.

Any time maintenance or forced outages require disassembly of cooling water piping to any component, it's important to circulate the water for some time after reassembly and then bleed any potential high points of air by whatever method is available.

As for "casues", well, it's generally disassembly and reassembly of cooling water piping that causes air to be present and to be trapped in the high points, followed by poor bleeding or no bleeding at all. The best bleeding is done multiple times.

I doubt that flame was lost because of lack of cooling to the flame detector(s). More likely the flame detector failed to detect the flame and erroneous 'Loss of Flame' trip occurred. Lack of cooling water to the flame detectors which are not present on every combustor would not cause the flame to be extinguished; maybe not "seen" by the flame detectors, but not snuffed out.

While it's possible that air could be leaking from either the atomizing air pre-cooler or the purge air cooler tubes/tube sheets, that could be easily checked just by opening the shell drain valves while the unit is not running and the cooling water system is under pressure.

Another possibility is that the cooling water system fill wasn't done properly, or the quality of the coolant isn't what it should be and it's degrading.
 
If the atomizing air pre cooler leakage that mean high pressure of the air will accumulate in the cooling line because of high different pressure. But in our case the air is more when the temperature of atmosphere is high otherwise there is no air.

Also if water system wasn't done properly, that mean it cannot water transfer to steam in low temperature if you think the cooling water is pure water but in fact water is mixed with ethylene glycol. Legs cooler give alarm when it reaches 72 Degree Celsius.
Purge air cooler is checked but there is no leakage.

Sometime we force two flame detectors to one to avoid trip because of flame lose. as you know if three flame lose either in primary or secondary that will make trip.
 
If there is a leak in the atomizing air pre-cooler then CPD, which is at a higher pressure than the cooling water system will leak into the cooling water system (when the unit is running) and it will find it's way to high points in the system and collect there, until it is vented.

Same for the purge air cooler; the air is at a higher pressure than the cooling water when the unit is running so it will find it's waht into the cooling water system and then migrate to high points in the piping until it's vented.

The only way I know of to check for leaks is to pressurize the cooling water system when the unit is not running (and CPD and AA pressures are zero) and open the shell drains to see if any coolant (water/glycol mix) drains out of the cooler.

I don't understand the "... air is more when the temperature of atmosphere is high otherwise there is no air...." statement.

There are usually isolation valves for the cooling water supply to the flame detectors to allow removal of the flame detectors during maintenance without draining the entire cooling water system. When the piping is reassembled, there WILL be air in the piping and so when the isolation valves are opened it's necessary to vent the tubing to and from the flame detector cooling coils. And the only way to do this on most units is to loosen the compression (Swagelok) fittings while the system is under pressure, and this should be done at least twice with as much time in between with pressure and flow in the cooling water system.

I've never heard of flame detectors failing to detect flame because they get overheated and then work again after they cool. When I've seen water-cooled flame detectors fail because of overheating, it was "fatal" and they never recovered.

A common mistake with water-cooled flame detectors is to have too much flow to the cooling coils. In humid environments or on units which use evaporative cooling or chillers the humidity in the compressor discharge air can condense on the lens of the flame detectors, preventing flame detection.

There are supposed to be valves (usually three-way valves) at each flame detector which can be "throttled" to prevent over-cooling and moisture condensation on the flame detector lens. It's not a very scientific means of preventing moisture condensation, but it's about all there is. A lot of sites use non-contact infrared heat detectors to try to prevent too much cooling water flow, or too little.

The 'throttling valves' at the flame detector cooling coils should be shown on the Cooling Water Piping Schematic (P&ID) and there should be notes on the drawing to indicate the valves need to be throttled.

What machine is this that has temperature sensors on the turbine support leg cooling water jackets?
 
I don't understand the "... air is more when the temperature of atmosphere is high otherwise there is no air...." statement.

That mean when ambient Temp is higher than 40 Celsius centigrade then more air accumulate in cooling water piping. But when the ambient Temp is less than 40 Celsius centigrade then little air accumulates in the piping.
 
It's very difficult for me to understand how ambient temperature could affect the amount of air which is accumulating in the cooling water piping.

You haven't indicated if there are throttling valves at the cooling coils for the flame detectors and if they have been properly adjusted.

You also haven't told us about the ambient conditions at your site (except that it's pretty hot!). Is it humid? Does the unit have evaporative coolers, or foggers on the inlet?

Again, a very common problem is that there's too much cooling water flow to the flame detectors, causing moisture to condense on the lens and disrupt the ability of the detector to see the flame. Are you sure that it's a lack of cooling water flow because of air that is causing the flame detectors to not properly detect flame?

Sometimes people will use non-contact infrared temperature sensors to obtain temperatures. These can be misleading sometimes, especially in a hot turbine compartment. Some detectors are better than others. The best way to obtain temperatures is to attach thermocouples to the same places on the flame detectors and/or the cooling water coils on the flame detectors and then run the leads out of the turbine compartment. One can then use a monitor or even a meter with a temperature converter to read the temperatures periodically and plot the temperatures versus ambient temperature and load and time.

A "high point" in piping occurs whenever the piping passes through a low point, up to a high point, and then back down. It's not *the* single highest point in the piping system. There can be multiple high points in piping systems which can trap air.

If you are certain that air is accumulating in piping, then you need to study *BOTH* the Piping Schematics (P&IDs) for the Cooling Water System and the actual, physical piping and try to determine the closest possible source of air, because the Cooling Water System, when operating, is under pressure, and so ambient air can't "leak" into the Cooling Water System. It has to come from an area of higher pressure. Or, it might be coming from a faulty pump, but that should not be happening if the Cooling Water System has a properly positioned and vented expansion tank. If the Cooling Water System is one of the "skid-mounted" system usually provided by GE or one of its packagers, then any air which might be introduced by the pumps should be vented in the expansion tank above the cooler modules.

Please provide some feedback and information, as many people read these posts and learn from them.
 
It's very difficult for me to understand how ambient temperature could affect the amount of air which is accumulating in the cooling water piping. Fin fan cooler is controlled by the read of inlet and outlet water temperature of the module.

These values can be controlled with temperature gauges installed on inlet and outlet module’s water piping (scale 0 to 100 °C). These temperatures are depending of the ambient temperature and the gas turbine load. For example for a maximal ambient temperature of 50°C , water at the inlet of the module must be at 67.3°C ( theoretical value ) and the outlet water module temperature with 4 fans per module running must be 61 °C ( theoretical value). An important divergence between water temperature measured on site and the theoretical values will indicate a problem of thermal exchange.

You haven't indicated if there are throttling valves at the cooling coils for the flame detectors and if they have been properly adjusted.
We are cleaning Flame detectors periodically from dust and also water cooling line vented at the same time. But when the flame detector failed to detect the flame, we are vented then the flame occurs at the running condition.

You also haven't told us about the ambient conditions at your site (except that it's pretty hot!). Is it humid? Does the unit have evaporative coolers or foggers on the inlet? The ambient temperature change from this range 8°C to 51°C.Ambient humidity change from 10% to 95 %. There is fin fan cooler .we have four module. Each module is made up of a supporting frame, one heat exchanger, four plenum and four fans.

Again, a very common problem is that there's too much cooling water flow to the flame detectors, causing moisture to condense on the lens and disrupt the ability of the detector to see the flame. Are you sure that it's a lack of cooling water flow because of air that is causing the flame detectors to not properly detect flame? I am sure that’s flame detectors failed to detect the flame because the air accumulate in the cooling line. When the cooling line vented then the flame appeared.
 
Top