Lube Oil Cooler, unecessary?

[nucleareng78]

We are installing a Pall lube oil purification skid that will remove particulates, free water, and dissolved water. Each tank is around 600 gallons and our skids will operate as kidney loops.

The existing system uses a filtration skid (only some free water removal) and a oil cooler that uses a fan to cool it down. Question is, since it's simply a kidney loop, is it necessary to have an oil cooler? The new skid has a built in heater that can be set to kick in when needed. The whole skid will shut down when the temperature of the oil gets too high. Flow rate is around 5 gpm.

Just wanted some operating experience regarding using these type of air oil coolers.

 

[tbuelna]

nucleareng78,

With a flow rate of 5 gpm and a fluid volume of 600 gallons, your system turnover rate is 120 minutes (2 hours). Whether or not you need a heat exchanger will be determined by the heat rejected into the oil and what temperature you need to maintain the oil at. With such a slow turnover rate, the answer is probably no. But you still should run some calculations to verify.

Hope that helps.
Terry

 

[nucleareng78]

I need to compare it with the flow rate of the lube oil pumps that take oil.from the tank to the pump bearings. Since the oil is spread out I er such a large area I can probably justify not needing a cooler.

When you say doing calculations, if I need the temperature of the oil coming.back to the tank could I do a heat sink calculation to find out what the average temp of the oil may be at any given point?

 

[rakuday]

nucleareng78,

Heat load will have to be calculated based on the heat removed from the bearings and gears (if any). This you can easily get from the bearing design/designers. The operating oil temperature and the heat removed will together have to be used for the heat load calculations. In all cases oil temperature is very critical for bearing operations.

You will have to calculate the area required to reduce the oil temperaure for operating conditions. This area will determine if you nee an oil cooler. If your reservoir size is too huge, such that the oil can cool inside (based on surfae area) then you might not need an oil cooler, else you would need one.

To start with you can use the below equations:
Q=m.cp.(T2-T1)
Q=U.A.LMTD

Q = Heat load
m = mass flow rate
cp = specific heat
T2 = Oil inlet temp
T1 = Oil outlet temp
U = Overall heat transfer coefficient
A = Area
LMTD = Log mean temp difference

Hope this helps.

 

[tbuelna]

I need to compare it with the flow rate of the lube oil pumps that take oil.from the tank to the pump bearings. Since the oil is spread out I er such a large area I can probably justify not needing a cooler.

When you say doing calculations, if I need the temperature of the oil coming.back to the tank could I do a heat sink calculation to find out what the average temp of the oil may be at any given point?

First of all, if there are bearings/gears involved, the temperature of the incoming oil flow, the oil mass flow rate, and the temperature rise of the oil, are what matters. With a very low oil flow turnover rate in your system, it is quite likely that there will be enough ambient convective heat transfer in the reservoir walls, piping, etc. such that a supplementary cooler will not be necessary.

 

[hydtools]

Heat transfer from the reservoir walls does depend on the oil contacting the walls. Heat transfer through oil is relatively poor. Time is the parameter. How long will it take to raise the bulk temperature to the upper limit compared to the duty time? How long will it take for the hot oil to cool down to a useful temperature compared to the machine idle time? The 600 gallons of oil has significant thermal inertia. How long does it take to heat 600 gallons to the minimum useful temperature with the built-in heater?

Ted

 

[CH5OH]

since you running back en from a storage tank, adding a cooler to the filtration unit makes it heat neutral.I am not familiar with your type of filtration unit, does it boil out water under vacuum (vacuum chamber/ejector)?if so you need to heat up the oil to about 80degC, storing oil this hot would degrade the oil quality, as well as making the oil not ready for use (typical requires 40degC or so)

 

[nucleareng78]

it is a vacuum dehydrator with a heater. The 6kW heater can be set to kick in at a certain temperature. During normal plant operation the heater may not need to come on due to the temperature of the lube oil after going through the bearings and controls.

I believe the tank is large enough to that we do not need a cooler. The entire skid is set to shutdown is the incoming oil is over 180degF, which is in the PLC's program.

 

[byrdj]

not sure of your application, but thinking about steam turbine system,
the return to main tank, drain from bearings can run very close to, if not above 180F (fossil). I don't recall average bearing return on Nuke, but would expect it around 170F. As a "kidney" unit, I would think the supply would be just from the reservior. Thus depending on the optimum temperture for the Pall unit, a cooler might be used occsional.

If the supply to the Pall will be from bearing header, then that should be controlled at 110 to 120F

 

[nucleareng78]

Since it is a kindey loop, wouldn't the oil have a chance to cool down enough?

The suction of the skid will take oil towards the bottom of the lube oil tank so the hotter oil will reside on the top part of the tank, from my understanding. I can't imagine oil from the feedwater pump and feedwater pump turbine being anything over 120F....I think

 

[byrdj]

Just looked at PI from a Nuke. the FPT and Main unit bearing drains were running 140 to 150F.

I just realized I was thinking bearing metal temps in the post above, SORRY.

 

[nucleareng78]

I need to locate any temperature indicators on the return line of the oil and maybe this will tell me the setpoint. Surely if your line oil is running too hot an alarm would trigger indicating a problem with your pump or turbine

 

[nucleareng78]

*lube oil

The feedwater pumps operate at 13000 gpm and 2280 ft of head. If I know the RPM I may be able to better analyze just how hot that lube oil is coming back into the tank. When it comes back into the tank, it's actually on the opposite side of where our new skid is taking in the oil.

 

[hydtools]

nucleareng78, measure intake and return temperature of the kidney loop. This will give you an indication of how much heat the loop adds. Then decide whether or not you need a cooler in the loop so as to have little or no affect on the lube system temperature.

Ted

 

[nucleareng78]

hydtools, good idea. The PLC on the skid can let you adjust how high to heat the lube oil up to (if wanted). I may mention in the operating procedure to adjust this parameter to desired temperature.

 

[tbuelna]

nucleareng78,

Using lube oil impingement to cool rotating components like gears and bearings is a very inefficient process. This is mostly due to the extremely brief period of time the oil is in intimate contact with the gear or bearing surfaces. The oil is squirted on and immediately flung off. The temperature rise in the oil in these situations is usually only about 30 to 40degF at best. Due to this inefficient heat transfer process and the low specific heat of lube oil, it can take a surprisingly large mass flow of lube oil to adequately cool components like gears and bearings.

You should also remember that lube oil cooling systems are usually designed to deal with some potential combined set of worst case operating conditions. As a result, the system is usually way over-designed for normal operating conditions. While it's a good idea to collect some oil in/out temperature data from a similar system operating at nominal conditions, just make sure you apply a reasonable factor to your calculations to allow for extreme cases before deciding whether or not a cooler is really necessary.

Good luck.
Terry