Air Cooled vs Water Cooled 9

[SparksZA]

As part of an EPC project we need to install some motors in the 3MW range and have requested air-cooled motors. The possible suppliers (2) have both indicated that water cooled motors will be more suitable for our application.

1. Does the water-cooled motors carry a bigger "risk" due to the possibility of cooling water getting into the windings?
2. Is it only a comercial issue with the water-cooled motors being cheaper than the air-cooled motors in these motor sizes?

Thanks

 

[asymptote]

The water jacket provides very effective sound attunuation which might be an advantage in some applications.

 

[Skogsgurra]

I have never seen a water cooled motor where there is an actual risk that water will get into the winding. The inner workings of such motors are usually very much the same as for an air cooled motor.

I would say that the risk of geting water in the windings is the same for both motor types. But then we are talking about water coming from the outside - like dripping water, flooding and other accidents.

Gunnar Englund

http://www.gke.org

 

[UKpete]

I haven't experience of large machines with water cooling but I have worked with high power PM machines using water and oil; I have no problem with water - in fact they tend to be cleaner internally.

The power density of a water cooled motor is significantly higher than that for air cooled. This is because the current density in the winding can be higher so the copper smaller in section, this allows some scaling down in diameter of the stator and hence of the whole machine - I'm not suprised that the suppliers want to offer it.

But of course, your decision must be based on how much trust you have in your supplier to get it right.

 

[SparksZA]

Thanks for the replies

The suppliers are both reputable (ABB and Siemens) so I do not expect any problems.

I have also never heard of a case where cooling water actually gets into contact with the motor internals. My management however is concerned that water cooled motors carries a bigger risk, due to close proximity of the cooling water to the motor itself. I have requested detailed drawings to determine the actual cooling design.

1. Does anybody has experience of water getting to the windings?
2. Is this perceived risk quantifiable?

Thanks

 

[edison123]

sparks,

In my twenty years repairing watter cooled MV motors and generators, I have not seen any case of cooling water entering into the winding. The coolers will have drain plugs for any tube leaks.

"Most people stop working when they find a job"

 

[vanstoja]

Water cooled motors are either of the "wet-winding" or "canned" design configuration both of which are designed to withstand the water environment with the former type probably more likely to be employed for power ratings in the megawatt range. Canned motor designs are generally limited to ratings of several thousand kilowatts due to structural integrity and power loss limitations of thin-walled stator cans employed and have motor efficiencies about 10 to 20 points lower than air-cooled motors of the same power rating. Generally, water-cooled motors are more expensive than air-cooled motors of the same power rating because of the added complication of the hardware needed to prevent the ingress of water to the windings.

 

[eglaude]

SparksZA,
I have worked with water cooled motors and MV drives for many years. My latest install was ~7MW, and looking back, we should have designed more than just the bare minimum for the cooling loop for better reliability. A flowmeter and circulation pump redundancy would keep you going. I never had any problems with leaks in the motor itself. The bean counters got to me on this project. Drives and exterior pipe systems are another story.

Keep your motor cool and you will have years of very reliable service. Beware of the power requirements of the driven equipmentthough. High power motors are usually SF 1.0 and are purpose built. If you push it too hard, for too long, you will need to monitor your stator RTD's closely and take appropriate action to save the motor.

 

[raisinbran]

SparksZA,
We use a variety of water-cooled motors in the underground coal mining industry, primarily to take advantage of the power density as UKPete mentioned. The water jackets that we use are essentially "added on" to the motor housing, so cooling water getting into the motor is generally not a problem. The motors are also explosion proof design. The largest that I have seen is 1650 HP.

The water jackets have a pressure rating that cannot be exceeded (we use 1000 psi testing), and the motors cannot be left outdoors with water in the jackets when it gets cold outside. You can limit the pressure with a relief valve, and we store our motors with anti-freeze in the jackets when they must be stored for prolonged periods.

A possible concern (in our industry) is impurities in the water collecting in the cooling jackets and plugging them up over time. We try to filter the incoming water, but it is not always effective. You may be talking about a closed loop system, but we use the water that exits the motor to cool our gearcases, and then use it for dust suppression.
Good luck,
Raisinbran

 

[rmgarnett]

SparksZA,

What are the ambient temperature requirements?
What is the operating speed?
Is this being driven from a variable speed drive?
Is water readly available?
What are the noise requirements?

The above factors will dictate the suitability of the cooling method.

RMG.

 

[aolalde]

SparksZA

I guess there are no 100% perfect machines. I have dealt with a Water Cooled Generator that leaked water and the winding got grounded.
That machine does not have a jacket water cooled, but four heat exchangers air-water that resembled a car radiator, assembled on the machine top. The inner cooling circuit handles air in contact with the machine windings and laminations, circulating in a closed circuit. The heat exchangers cool down that hot air with circulating water. The main factor for the leak to develop was corrosion due to water contamination.

The question you should answer is. Do I have available enough water with the required quality to cool the motor? that water will add to the energy operational cost.

 

[SparksZA]

rmgarnett

What are the ambient temperature requirements? 5 to 45 degC
What is the operating speed? 1800 rpm
Is this being driven from a variable speed drive? No
Is water readily available? Yes, soft water (treated, closed loop)
What are the noise requirements? 85 dB measured one meter from source


aolalde

I am still awaiting the design, but it should be similar to what you have described.

From the responses I have received it seems that there are no "additional safety risk" when water and air-cooled motors are compared, providing that the cooling water quality is good enough not to cause corrosion problems. Even with bad water it seems that the "risk" is increased only marginally due to a "general design" that prevents the cooling water to come in contact with the motor internals.

Thanks for all the posts

 

[sms]

The only risk in my mind is freezing the water jackets and bursting them. I did have a motor of this size get water in it, but not from the cooling system rather from drenching rain along with high winds. Made a lovely spark......and died......

-The future's so bright I gotta wear shades!

 

[rmgarnett]

What are the ambient temperature requirements? 5 to 45 degC
What is the operating speed? 1800 rpm
Is this being driven from a variable speed drive? No
Is water readily available? Yes, soft water (treated, closed loop)
What are the noise requirements? 85 dB measured one meter from source

Ambient temperature requirements will dictate the frame for the machine, if you are increasing above the standard 40degC rating then it is probable you will have an increase in motor frame size i.e. increase in cost.

Water cooling the motor is a far more efficient way of cooling than air, and could possibly provide a reduction in motor frame size. However, if the closed-loop water cooling system is via an water-to-air heat exchanger, where the ambient temperature is 45degC, the motor will be under the same constraints as that of air cooled. i.e. increase in motor frame, increase in cost. I assume that the water is chilled, hence there will be an advantage in going water-cooled.

If the motor is installed indoors then the cost of conditioning the air for that room will have to be taken into consideration, it will be generally cheeper to exhaust the heat losses outside. i.e. reducing the requirements of the Air Handling Unit for the room. In addition to this, if an air-cooled motor is located indoors, there is a risk of the exhausted air being drawn back in by the cooling fan, only exacerbating the problem of cooling.

In summary, UKPete is correct. The main reason why they have both quoted water-cooled is probably cost, but only due to the ambient temperature requirements of 45degC. (It would be interesting to know what the temperature of the cooling water is.) A 3MW motor with an eff. of 96% would mean losses in excess of 100kW, thats a lot of losses to deal with, hence another reason why manufacturers will quote water-cooled at this power rating.

Motors cooled with a water system who's temperature is considerably lower than the ambient air, will probably mean a smaller frame motor than that of an air cooled solution. Don't be put off by this method of cooling, water cooling is a very common practice, I am positive that both Siemens and ABB could provide you with an extensive reference list if you requested one.

************
The VSD is of no consequence here as it is a fixed speed application.
The noise level should not be an issue with an air-cooled motor at a speed of 1800 rpm.
************

I hope this was of some help to why they may have quoted water rather than air.

RMG

 

[UKpete]

rmgarnett, I was interested to read your comments. The machines I was familiar with were on CHP systems and used the return water from the building heating system for cooling - this could actually be at up to 85°C. The heat transfer coefficient of water was about 3500W/m²K compared with only about 100W/m²K for air, so even with the water somewhat hotter than ambient air, the winding would run cooler and the machine designed smaller.

Of course the same amount of heat has still to be shifted to ambient, but water cooling is much better at removing heat from the critical area where the area available for heat removal is small, i.e. the stator winding, allowing the machine size to be reduced.

 

[rmgarnett]

UKPete,

I have assumed that the motors have been rated for class B temp rise (90degC), with class F insulation.(115degC).

Motors are generally designed Class B temp rise, Class F insulation and for 40degC ambient, if they have to operate in temperatures above these conditions considerations have to be made within the motor design. This is pretty much a de facto standard; water inlet temperature for the cooling should be above 5degC, yet below 25degC.

The motors you mention must have a special design, or temprise must be different to the classes above. As a secondary point it is also worth considering that it is generally not possible to reach the hottest point of the winding.

Anyway, the reason for my post was to demontrate why water may be cheaper that air. It is very probable that an air cooled machine will require a larger frame, whereas a machine that is water cooled may be able to be supplied in a smaller frame. However, you must remember that the additional cost of the heat exhanger will have to be ballanced against the cost of the larger frame machine. At this power rating water-cooling should be more cost effective.

RMG

 

[magneticted]

Just to add a comment, for added comfort make sure a leakage detector is supplied with the machine, if you go this route. You could also ask the motor manufacturers for a reference list of water cooled machines supplied.

 

[UKpete]

rmgarnett - yes you are right, it was a special machine with in fact a full class H temperature rise and class C copper enamel, so it didn't represent what ABB or Siemens might be doing.

I'm not familiar with the specifics of industrial water-cooled machines; I only really wanted to stress that water has a heat transfer coefficient many times that of air, area for area, so that it is viable even without chilling (quite expensive relative to a water to ambient air heat exchanger I imagine). For example, it's considered the best option for a car engine.

SparksZA: the reason high power motors are a different proposition is that as motor dimensions are scaled up the problem of cooling gets worse (power and therefore losses going up with the cube of dimension i.e. D²L, whilst dissipative area for removing heat only going up with the square of dimension) hence the move towards water cooling.