Cooling Water System Expansion Tank


Thread Starter

Mustafa Jamal

We have gas turbine frame9E. we faced problem in cooling water system. the problem that during operation expansion tank level for example 80%. and when stop GT and GT became on cooldown, the expansion tank level reduce about 70%. but when start up GT, the level return same. we try filling tank when GT on cooldown, but when start GT expansion tank became over flow for all water we add. We checking piping for system but not found any leakage.


The temp. of legs is different between right side and left side. and when we make vent for one side, we found air inside system and temp. became equal. but after few hour the difference return.

I am unable to understand the meaning of expansion tank. Is it is cooling tower sump?

seems not from your post.

Can you provide some more details on it?

you can give me some photographs ([email protected])
Mustafa Jamal,

When fluids (and most metals and other materials) are exposed to heat they tend to expand (as the molecules grow further apart). This is the entire purpose of the expansion tank of any cooling water system: to deal with the expansion of the fluid without overflow when exposed to heat while running, and to be able to visually ascertain the level in the system at any time--while running or while stopped.

The expansion tank is not supposed to be filled to FULL when cold; in fact, if ever, it should only be filled to FULL when the system is running at Base Load in steady-state condition, when everything is good and hot and at stable temperatures. Even then, it's still not recommended to fill the expansion tank to FULL, because should there be some cause for the cooling water temperature to rise even further the expansion tank will overflow. The expansion tank is sized such that it should be possible to see a level at all times--when the unit is running, when it's starting, when it's stopping, when it's stopped and "cold."

And, just as when the fluid expands when it's heated, it contracts when it's cooled/cools, and so the level in the expansion tank will fall.

This is all normal and expected--unless there's more to the story that you haven't told us. Most every automobile with a water-cooled internal combustion engine has an expansion tank these days. Because the coolant (water and any enhancer added to it) will expand as it's heated, and will flow out of the radiator/closed system. But, if it flows into an expansion tank, then when the engine is off and cools, the coolant will be drawn back into the radiator/closed system--and not lost to atmosphere/ground.

Even the air in vehicle tires expands when the tires are rotating as the vehicle is driven. That's why it's recommended to check tire pressures when the tire is cold, when a reliable measurement of pressure can be taken. If one adjusts tire pressure when the tire/air is "cold" (before the vehicle is driven), and then checks the tire pressure after a long drive (when the tire is subject to normal operation) the air pressure inside the tire will be higher (presuming no leaks), and if one then vents air pressure from the tire to make it equal to rated, and then checks the tire air pressure again after the vehicle has stopped and the tires have cooled, the tire air pressure will be less than rated.

What you're observing is basic physics--and is the purpose of using an expansion tank on the turbine-generator Cooling Water System. It's not designed to kept full--it's designed to allow for expansion without loss, and to be able to view the level of coolant in the system at all times (while running and while cool when not running).

Hope this helps!
Mustafa Jamal,

In trying to understand other possible reasons for the questions, the fluid in the expansion tank isn't actively involved in cooling; it's just a storage tank for being able to see the level in the cooling water system both when the turbine is running, and when it's not (when it's cooled). The fluid in the expansion tank isn't being continually circulated when the system is running. (Review the Cooling Water System P&ID for details. P&IDs are excellent aids in understanding systems and operation. Many of the Operations- and Service Manuals provided with GE-design heavy duty gas turbines also have individual sections commonly referred to as 'System Descriptions' which have written, albeit brief, descriptions of systems and components and intended operation.)

If the reason for keeping the expansion tank full is to assist with cooling, that's not a valid reason for doing so. If the cooling water system is unable to maintain L.O. temperature and/or AA Compressor inlet temperature, it's most likely not the level in the Cooling Water expansion tank that's the problem, unless there is no level visible in the expansion tank sight glass, that is.

Again, the main purpose of the expansion tank is to provide a method for visually seeing a level in the Cooling Water system at all times. It's not an active part of the Cooling Water system; it's just a storage vessel which can accommodate changes in level during normal operation without any overflow on to the ground or into a drains tank (which must eventually be pumped out and disposed of). As such, it also prevents continual filling/draining of the system, reducing maintenance activities. It helps to ensure that the heat exchangers and the entire Cooling Water system are full during normal operation--even if the expansion tank isn't exactly full.

Another purpose of the expansion tank is to allow a high-point for air to exit the Cooling Water system. Many times, during filling of the Cooling Water system, air can be trapped in various parts of the piping and components, and eventually this air usually finds it way to the highest point in the system--which is the expansion tank which is vented to atmosphere.

By maintaining a visual level in the expansion tank at all times (even when it's not FULL), one can monitor drops in the level and add fluid before the level gets too low (while most expansion tanks have low level switches to provide an alarm, many of the switches, because they are exposed to the ambient environment, have failed and not been replaced over the years). If it become necessary to frequently fill the expansion tank, that's an indication of a leak of coolant/fluid. Or, if the fluid continues to rise and overflow, that's a possible indication of a leak in the L.O. Coolers (L.O. is at a higher pressure than Cooling Water, so that water doesn't (shouldn't) leak into the L.O. system, but rather L.O. would leak into the Cooling Water system).

The Cooling Water expansion tank serves many different purposes, but most importantly that of maintaining a visual indication of the level of fluid in the Cooling Water system.

If there's a problem with L.O. and/or AA Compressor Inlet temperatures (or any other temperature maintained by the Cooling Water System), please describe the problem and we may be able to help--but do be prepared to provide more information when requested to get the most concise response. While every GE-design heavy duty gas turbine compresses air, burns fuel, and exhausts they can all do so with some very unique--and important--differences which can be very important to troubleshooting.
Dear Mustafa Jamal,

To complete CSA comments about Expansion Tank and vehicles...
Most new cars and trucks are delivered with pressurized water cooling circuit (with expansion tank, not atmospheric or closed system). To increase the <b>thermal efficiency</b> of an internal combustion engine you need to increase its internal temperature. This internal temperature depends of the temperature of your coolant. This is why the coolant (water + some additives) is kept at <b>higher</b> than atmospheric pressure to increase its boiling point. <b>A pressure relieve valve</b> is usually incorporate to the radiator's fill cap.

I'm not familiar with Frame9 but expansion tank means pressure relief valve. I think you need first to identify and check this relief valve and if this valve is working properly that means you may have a located overheating (or cavitation) somewhere on your cooling system. This relief valve should exist on the machine P&ID?!

Another thing; Follow OEM instructions when filling your tank and checking the water level. On vehicle; it's <b>forbidden</b> to open the cap of a hot radiator (ejection of dangerous hot water and steam)

Hope this is helpful

Many GE-design heavy duty gas turbines were/are supplied with on- or off-base cooling water systems. Earlier packages had the water-to-air heat exchangers mounted above the Accessory Compartment, some even had fans to move the air over the hear exchanger fins driven mechanically from the Accessory Gear box using long, articulated rods. Later designs used AC induction motor-driven fans to circulate the air. These systems were considered to be "on-base", as they were attached to the compartment enclosure, and everything was mounted on the same concrete foundation ("base"). The Cooling Water pump was also Accessory Gear-driven.

Later and more recent designs for plants which did not have cooling towers or other cooling water systems had a separate foundation on which electric motor-driven Cooling Water pumps and a water-to-air heat exchanger with electric motor-driven fans were mounted. This was called the "off-base" design.

In both of these systems, the Cooling Water system expansion tanks were NOT pressurized, but were open to atmosphere. (Yes; there is a pump, or pumps, to circulate the water under pressure, but the system was open to atmospheric pressure. No relief valve.)

Most of the turbines supplied with "self-contained" Cooling Water systems like the ones described above are for utility peaking applications, which do not have any other need for Cooling Water. You seem to have experience in combined cycle applications, where there is usually a steam turbine and condenser, gland seal condensers, and steam turbine L.O. cooling requirements--which means there is a very large need for Cooling Water in addition to that required by the gas turbine(s). In these types of applications, the gas turbine cooling is but a fraction of the overall load of the Cooling Water system, and can be easily supplied without the need for separate heat exchangers and pumps solely for the gas turbine cooling requirements. In this case, one does look at the cooling tower sump level to determine if the system needs additional fluid or not, and, generally, it's also open to atmosphere (the cooling towers and sump).

The only point in bringing up the expansion tanks on vehicles (which are ALSO open to atmosphere) was to try to draw a parallel with other equipment which employ expansion tanks for similar reasons: to check system fluid level, and to contain overflow. Again, on the on-/off-base, self-contained Cooling Water systems, there is no relief valve on the expansion tank, so none would be present on the P&ID; it's open to atmosphere, just like the expansion tanks on cars and trucks. Anyone can open the lid of a car or truck expansion tank (not the radiator--but the expansion tank) at any time, realizing the fluid inside may be very hot. The "bottom" of the gas turbine Cooling Water system expansion tank is connected to the Cooling Water system, just like a vehicle's expansion tank, so that fluid can flow back and forth between the systems as needed. The biggest reason for pressurizing a Cooling Water system is to reduce the size of the system, something important for cars and trucks, but not so important for gas turbines in utility peaking applications.

While many GE-design gas turbines are very similar and operate on the same basic principle of suck-squeeze-burn-blow, they do so with very different auxiliaries, especially when used in different types of applications. While it's common for many to think all GE-design heavy duty gas turbines are the same, they are not. There are many types of automobiles--but the primary goal is the same: move people and/or goods. And that can be done in many different ways, with many different designs and auxiliaries. Some automobiles have air-cooled engines and don't have cooling water systems at all. Gas turbines are the same--even those made by the same manufacturer, or under license from the same manufacturer. (Some gas turbines are used for driving compressors and/or pumps--even GE-design heavy duty gas turbines; not all are used for electrical power generation.)
Dear Mustafa;

>The temp. of legs is different between right side and left
>side. and when we make vent for one side, <b>we found air
>inside system</b> and temp. became equal. but after few hour the
>difference return.

Looking to my documents I found this: The MS9001E HDGT cooling water system is designed to accommodate the heat dissipation requirements of the lubricating oil system and the turbine supports.

And the expansion tank is there to ensure minimal pressure at water pump suction and to compensate water volume variations due to thermal dilatations and eventual leakage.

I think you have the solution to your problem: <b>AIR in COOLING SYSTEM.</b>
Air pockets or air bubbles are very harmful to the cooling circuit (cavitation phenomena) especially if it's a closed cooling system.

May be you have a leak and air is going inside your cooling devices or may be the GT legs are too hot and water is boiling at that point?!
On Frame 9, correct me if I'm wrong, legs or turbine supports are cooled to avoid any misalignment. Are you experiencing any kind of vibration?

One more thing; if your expansion tank doesn't provide enough pressure at the water pump section cavitation may appears

Dear Mustafa,

are you facing a drop in WC pump discharge pressure? if so, kindly clean the strainers at pumps suction then vent the water system with both WC-pumps off and let us know the result please.

Mustafa Jamal,

maythem and bkarim55 bring up good points--the suction strainers of Cooling Water pumps can often become clogged and restrict the flow of Cooling Water. Low flow can contribute to the collection of air in various high points in the Cooling Water system. (High points are defined as points in the system above the "inlet" and "outlet" points for the entry and exit of Cooling Water. For example, the inlet and out of Cooling Water to the cooling water jackets of the aft turbine support legs are usually below the highest point of the jacket--meaning that air can be trapped above the exit and must be removed by using some kind of air bleed valve. Any section of the Cooling Water system can have high points, and even multiple high points--especially the flame detector cooling coils and associated tubing. There is usually one place the flame detector cooling water system is supplied from and it all goes back to a single "exit"/return to the Cooling Water system. Between the "entry" and "exit" there are usually several high points in the piping/tubing where air can and does collect and must be manually vented from. High points don't always show on P&ID drawings; they are typically just schematic representations of systems, and do not reflect all of the piping/tubing and elbows and high- and low-points.)

And, air in aft turbine support legs is the usual cause for discrepancies in support leg temperatures. And, the usual cause of air in aft turbine support legs is poor filling and air-bleeding procedures after a maintenance outage. It's very important to bleed all or as much air out of all high points of the Cooling Water system after filling/re-filling the system. And, while technically that's physically impossible to do, the circulation of Cooling Water usually moves any air in the system eventually to high points in the system, so it's customary to periodically bleed air out of high points between maintenance outages, and especially for some period of time immediately after a maintenance outage.

The aft turbine support legs usually have manual vents for removing air, and I've been told that lately GE-Belfort has been providing some kind of "automatic" air-bleed vent valve on the aft turbine support legs (though they don't seem to work well). If the unit has DLN-I combustors with water-cooled flame detectors, many of the newer systems don't have air-bleed valves in the high points of the tubing, so it's necessary to loosen the compression fittings to bleed air out of the tubing of those systems (and that's only usually possible when the turbine is not running).

So, if the problem is uneven aft turbine support leg temperatures, the temperature gauges are usually abused during maintenance outages and can display differential temperatures when none really exist (so another means of measuring temperatures must be used to verify), and air can move around the Cooling Water System and build up over time in the tops of the cooling jackets of the aft turbine support legs.

As mentioned, the aft turbine support legs are water-cooled to prevent excess growth in the support legs from causing misalignment and vibration between the turbine and generator. So, usually, if there is a genuine high differential in aft turbine support let temperatures there will also be a noticeable rise in #3 bearing and/or turbine-end generator bearing vibrations as the turbine is loaded and the leg temperature differential increases. This isn't always the case, but it is more common than not.

Lastly, the expansion tank--being vented to atmosphere--also serves as the place where air in the system can be "lost" to atmosphere, provided it is moved around and through the Cooling Water system by the Cooling Water Pump(s). And, the design of most Cooling Water systems is such that not all air can be dislodged from the multiple high points in the system simply by Cooling Water Pump pressure and flow. So, it's necessary to periodically manually vent the various high points of the Cooling Water system to ensure that air doesn't collect in them, which it can over time if not vented (manually, or automatically if the automatic vents provided work properly).

Again, it's pretty common for air not to be properly and fully vented (to the extent possible) after maintenance outages, and for air to collect in high points and need to be vented--perhaps several times. And, also, it's customary for high point vents to be checked periodically between maintenance outages to ensure as much air as possible is vented from the system as it naturally moves through the system. (By the way, this is also true of gas turbine cooling water systems which are supplied from large cooling towers/sumps--that air must be bled from the systems after maintenance outages when the Cooling Water piping has been drained, disassembled and reassembled.)

Hope this helps!
Dear CSA;

Thank you so much for the huge information you are providing.. I just would like to make a comment about vehicles expansion tank. Most of recent Europeans cars are delivered with pressurized cooling system that means coolant pressure is higher than atmospheric pressure. When engine is switched off and water pump stopped, coolant thermal expansion will happen causing a pressure build-up in the cooling system. To drop this pressure down, calibrated relief valve(s) is/are installed on the fill cap(s) of the expansion tank or/and radiator.

On the Web, I found this document :

Which is a Draft Technical Specs for GE Frame PG9171E and on its 62 page we can read:

"<b>Atmospheric Expansion Tank</b>
The atmospheric type expansion tank is installed at 6 meters height on its own steel structure support.

This 0.9 m3 capacity tank is fitted with one local level indicator combined with a low level switch for remote indication and a connection pipe for make-up.

<b>Pressurized Expansion Tank</b>
The expansion tank is installed at ground level. It is pressurized at 0.6 bar(g) with nitrogen or air.

This 0.75 m3 capacity tank is fitted with: One pressure relief valve on N2 (or air) side."

What's your opinion? Is this just a draft or this kind of design (Frame 9 with pressurized expansion tank) is available?

Thanks In advance and best regards

I've said it before, and I'll say it again: I haven't seen every GE-design Frame 9E heavy duty gas turbine that GE Belfort has come up with. They are known for stretching the envelope, and a pressurized Cooling Water system may be another of their we-will-because-we-can and we-have-a-better-idea "innovations."

One of the really good things (and there are a lot of them) about GE-design heavy duty gas turbines was always that they employed tried-and-proven concepts that generally made them simpler to operate and maintain. Belfort doesn't believe in simpler-is-better.

Can you <i>honestly</i> believe that pressurizing a Cooling Water system expansion tank--which means gauges, high- and low pressure switches and/or pressure transmitters, a relief valve, a pressure regulator and a source of nitrogen (if that's used for the system), plus additional control system I/O capability and associated application code ("logic" if that term can be used for over-complication) improves operation and maintenance and availability/reliability?

The origin and purpose of the document you cited is unknown. It appears to be a specification written for use in a request for a proposal for offers to supply a turbine-generator and auxiliary, and may have been (likely was) written by a third-party or A/E (Architect/Engineer) who included that as an option for an undisclosed (unknown) reason or purpose. There's no way to know from that document if a pressurized Cooling Water expansion tank was ever built, installed and successfully commissioned by GE or a packager of GE-design Frame 9E heavy duty gas turbines and auxiliaries.

The point, bkarim55, is: Even if such an expansion tank was installed, your reply mentioning a pressurized expansion tank assumes that all GE-design Frame 9E heavy duty gas turbines are so equipped. (And you cleverly--and conveniently--found a document on the World Wide Web of unknown origin and purpose that you are using to support your contention after the fact. Your original post posited that the original poster may have been referring to a cooling tower sump... ;) ) You didn't ask the original poster if his unit was so equipped or mention the possibility of other possible configurations and their potential issues, if any. Even if an expansion tank is pressurized, it still doesn't need to be continually full during operation and shutdown; the level is still going to change as the temperature of the water/coolant changes even if it's under pressure. And the level in the expansion tank--pressurized or not--isn't going to effect the aft turbine support leg temperatures, which wasn't clear from the original post and still hasn't been clarified by the original poster.

Not all GE-design Frame 9E heavy duty gas turbines are the same, as not all cars in Europe are the same. Even European vehicle manufacturers use different components and configurations on the same model of a vehicle, making troubleshooting very challenging. It's important to recognize the differences when posting and take them into account or ask for clarification (which we've never received from the original poster).

I have learned from your research that there are or may be configurations I have not seen, and may never see. But the physical principles and the the purpose of a Cooling Water expansion tank--pressurized or not--are still the same. In such cases, I (we) have to rely on more information and clarification from the original poster--which we oftentimes never receive in a timely fashion. This is all part and parcel of posting replies to It's not always easy, and it's sometimes frustrating, but it can be extremely rewarding and satisfying.

Let's agree that there may be pressurized Cooling Water expansion tanks but that the principles and purpose are the same: to provide level indication, to prevent overflow and to supply some "make-up" water/coolant as fluid temperature changes or leaks., as well as a way of venting any air from the Cooling Water system.

And leave it at that unless and until the original poster writes back wit more better information.

We've kicked thus poor horseofva topic just about to death and can leave it safely as if for the time being until such time as we may receive clarification from the original poster.

Thanks, we collaborate in similar projects in the future
Dear CSA,

I appreciate and its little search box :)) And I really appreciate your simplest way to describe events.. It will be a great pleasure to collaborate with you. Any time you're welcome

I wish Happy New Year to and hope you are enjoying yourself with your family.

Expansion tanks in both heating and chilled water systems (and domestic hot water heating) are used to limit the expansion of water during times of varying temperatures. This aids in the limiting of the pressure of the system and keeps things from over pressurizing and blowing off T&P valves.

As temperature increases, so does the specific density, and pressure, and thus the water expands. If it doesn’t have anywhere to expand (expansion tank), the pressure will increase proportionally.

Another thing to think about of a similar vein, but different, is in chilled water systems, it’s best to also install a buffer tank. This adds to the mass of the system to help reduce the effects of expansion, but more importantly to reduce the cycling of the chiller.
cooling water system consists of a refregerator system in which refrigerant becomes cool and hot subsequently in one comolete cylcle. the basics of refrigeration cycle is to adiabaiticaly compress the gas (refrigerant) so that its temperature increses higher than atmospheric temperature and loose heat energy in atmosphere and cools down nearly to atmosphere and it again gets cool below atomspheric temperature down by expansion process. so that expansiom tank is used which has throttle valve through which sudden exapnsion of gases occurs and gases becomes cool.
n circulating-cooling-water the water often has temperature difference that would cause the pressure of water in pipe unstable. For example, when water be heating the volume of cooling tower would increase that would have safe problem when bigger pressure on whole piping system. In this time, expansion tank be equipped to balance the water pressure in pipe. The operating of expansion tank is when water temperature getting cool or water pressure in pipe is getting lower the storage water of expansion tank would inject the water into pipe to keep water pressure stably which in pipe.