Excess Heat Generated By a Liquid Ring Pump

[Amused_1]

Hi all. First, some context.

I work for a small seed oil extraction plant where we cold press the seed then solvent extract the residual oil from the presscake. In the solvent extraction plant we use liquid ring pumps to pull the required vacuums. The Plant Vac Pump (PVP) is used in all operations and pulls the vacuum for the 1st effect, 2nd effect, the Oil Stripping Column (OSC) and the Mineral Oil System (MOS, also known as the Solvent Air Separator). This typically pulls a vacuum of around -600 millibar when distillation is in full swing.
We also have a High Vac Pump (HPV) that pulls a vacuum on a separate vessel called the High Vac Dryer at around -850 millibar, but this is used only when the finished oil requires a <1ppm residual solvent content.

Both pumps use water as a service liquid, which is typically around 20-30 Celsius.

Recently, when using just the PVP, we’ve noticed the top of the Vac Tank getting hot. The water doesn’t seem to increase much in temp, but the space above it (and the ceiling of the tank itself) feels like it could be around 50-60 Celsius. We’ve traced it back to the pump itself - the inlet pipe for the water is no warmer than usual, the vapour inlet that the pump is pulling on is cool as any air/vapour present has been pulled through a condenser, yet the outlet for the water is again around 50-60C.

We’ve gagged back the valve on the vapour inlet which has reduced the vacuum to between -500 and -550 and found this has reduced the temp on the outlet, but moving forward we’re not sure how to solve the issue.

Would anyone have any thoughts? I can provide more info if needed, but figure there was plenty above to chew on!

 

[Bystander]

"Drying oils?" As in linseed?

 

[Amused_1]

Yes, though we haven’t done linseed for years. We’re a toll extraction company for specialty seed oils

 

[ChemAir]

We had multiple liquid ring pumps for removing excess alcohol from plasticizers. All our liquid ring pumps were either 1) once through with water, so heat couldn't build up, 2) they had a captive circulation system with a water cooled heat exchanger in the circulating loop, or 3) a captive circulation system that had a large enough circulating reservoir that the pump never operated long enough to heat the tank. I think if you look at Nash-Gardner-Denver you'll find they will design a skid system that can pull vacuum, and cool the seal liquid. They are who I'd call for heat exchanger recommendations.

In my facility, once through water was for pulling vacuum on systems that didn't generate significant light end condensation. Circulating systems accumulate light ends, hydrocarbons, etc, that gradually affect vacuum performance and influence pump horsepower, and usually had some type of purge to a waste treatment system to clean the water back up, and recycle some of the product that made it into the vacuum system to a better location. My experience with keeping liquid (water) ring pump heat exchangers clean is that they require more than typical heat exchanger cleaning, but that is probably the specific service.

Be careful when your vacuum tank gets hot when it is light end contaminated. Some could be combustible, and the ignition risk can be high if consideration hasn't been made.

 

[Amused_1]

It’s difficult to explain the setup, but my drawing skills aren’t great so here goes...

The main parts of our vacuum system are:
The work tank - this is where all condensed water/solvent returns to from the plant. The solvent (Hexane) separates to the top and flows over a weir to the ‘clean’ side where it is reused. The water is pumped from the bottom up to the vac tank.

The vac tank - this is a reservoir of water that is connected to the liquid ring pumps. It also has a weir in case of overfilling, the other side of which is where the water that has been through the pump returns to before flowing through a cooler/condenser; if any solvent vapours are present with the used water they are pulled through the same cooler/condenser but pulled through a different vac system afterwards.

The vac pumps (PVP + HVP I mentioned earlier) - these pull the vacuum for the plant. The service water is pulled by the pump as opposed to being fed by any kind of water pump. The air/solvent vapour/water vapour they pull from the plant is passed through a condenser before reaching the pumps, in theory leaving non-condensables such as air to be the only thing passing through the pump.
I’m wondering if the pump isn’t pulling enough service water to have the cooling effect I would expect to see.

To try and nutshell the issue, my questions are-
Should the pump(s) be generating heat in the service water passing through if they were setup/working correctly?
If not, then could a lack of service water cause the heat?
Furthermore, is it feasible to have the water pump-fed to the liquid ring pump?

The heat itself isn’t a concern as the flashpoint of Hexane is far higher than the temps we’re talking about, but fair point.

 

[ChemAir]

I’m wondering if the pump isn’t pulling enough service water to have the cooling effect I would expect to see.

Be aware, that despite their ability to pull vacuum, liquid ring pumps aren't necessarily great liquid pumps. I recall a specific application where a heat exchanger in the recirculation system introduced enough pressure drop that required a second, liquid pump for the ring pump to operate properly. Again, Nash-Gardener-Denver is who I'd go to to discuss this. If yours are made by a different manufacturer, you should contact them to get this information.

Should the pump(s) be generating heat in the service water passing through if they were setup/working correctly?

Yes. It's a consequence of gas compression and work done in the pump. Too much water temperature is because water started out too hot, or water flow is low. Hotter water=less vacuum.

If not, then could a lack of service water cause the heat?

It doesn't cause the heat, the motor HP used in the pump does. Lower liquid flow means higher liquid outlet temp. If you continue to recycle the same water without removing this heat, the temperature will increase in the whole water system. In the scheme of things, this heat input isn't large. Often just convection around a tank, a little fresh water addition, and some evaporation, can be enough for smaller systems. A large pump and a small tank can require additional cooling.

Furthermore, is it feasible to have the water pump-fed to the liquid ring pump?

Yes, and possibly preferred in some installations, but you need to know what the flow requirements (temperature and pressure) are for the pump. If you overload the pump with liquid, it won't work like it should, either. This is where you should call the manufacturer to get minimum liquid flow required, cooling required, etc., for guidelines.

 

[Amused_1]

Thanks for the response.
Normally, with the used service water returning to the work tank before being pumped back up to the vac tank to service the liquid ring pump again, it mixes with enough cool water from the condenser vessels that it is a good operating temp when it gets back to the vac tank, so I don’t think there is an issue there.
The heat in itself isn’t an issue as such, I just believe it’s a symptom of something else.
Gagging back the vacuum side of the pump reduced the temp of the output water a lot, which is why I’m questioning the service water feed.
It’s a Travaini TRMB 40 - 150 http://tpusa.brinkster.net/_MasterDirectory/100747.000_TRM_40_Cut_Sheet.pdf
On the tech sheets it states an average liquid flow of 5GPM or 4.5Ltr/min, but we have nothing on our setup that would allow us to measure this.

 

[ChemAir]

Re-check this:

5GPM or 4.5Ltr/min

It’s a Travaini TRMB 40 - 150 http://tpusa.brinkster.net/_MasterDirectory/100747.000_TRM_40_Cut_Sheet.pdf

That one is not very big, only 5 hp. If all 5hp went to heat, it shouldn't raise 5gpm of 60-70 F water much more than five degrees F (if my back of the napkin math is correct). I'd ask the manufacturer what the actual expected rise is at the correct flow and your operating conditions.

Since you don't yet have a way to measure flow, I'd suggest a couple things
-Figure out a way to add a clamp on meter, install a temporary flow device, or empty the water tank and time the fill up to back calculate the flow. You can also direct the water to a drum or tote on a scale and get a stopwatch. The water flow rate should be measured while the pump is running at normal operating pressure/vacuum.
-Once you have the factory expected temperature rise, as a sanity check, get a good, accurate water temperature before and after the pump, and try to ballpark back figure how much water is actually getting to the pump.
Once you know the flow, you can go on to the next step.

 

[Amused_1]

Thank you very much. I’m currently on my fortnight off (the perk of shift work!) but when I get back I’ll have something to work with [emoji106]

 

[Asymptotic]

The vac tank - this is a reservoir of water that is connected to the liquid ring pumps. It also has a weir in case of overfilling, the other side of which is where the water that has been through the pump returns to before flowing through a cooler/condenser; if any solvent vapours are present with the used water they are pulled through the same cooler/condenser but pulled through a different vac system afterwards

For 'cooler/condenser', I'm envisioning a water-cooled heat exchanger. As exchange surfaces foul and/or passages become progressively more blocked the amount of heat that can be removed gradually worsens, eventually to the point where working fluid temperature goes too high. Could such an effect have any bearing on your problem?

 

[Amused_1]

For 'cooler/condenser', I'm envisioning a water-cooled heat exchanger. As exchange surfaces foul and/or passages become progressively more blocked the amount of heat that can be removed gradually worsens, eventually to the point where working fluid temperature goes too high. Could such an effect have any bearing on your problem?

In this case, no. It’s a bullet-type shell and tube heat exchanger, and while it could suffer from the problem you mentioned it seems to be working fine. We even increased the flow of cooling water through it just in case but I think it might have been unnecessary.

 

[Asymptotic]

In this case, no. It’s a bullet-type shell and tube heat exchanger, and while it could suffer from the problem you mentioned it seems to be working fine. We even increased the flow of cooling water through it just in case but I think it might have been unnecessary.

If tube surfaces are fouled (provided the usual case that water flows through the tubes) increased coolant flow will not yield the expected result.

A way to find out is install a flow meter in the inlet pipe, and accurate thermometers (or another sort of temperature sensor) in the coolant inlet and return pipes. Delta T and mass flow rate are used to calculate heat removal. You'll want to find the relationship stated in SI units, but the one burned into my memory is BTUH = GPM * 500 * delta T (°F). The 500 value comes from 8.34 pounds * 60 minutes per hour. Specific Heat and Specific Gravity factors are both 1.0 for water and can be ignored.

If heat removal doesn't increase by very much when water flow is increased, transfer between the tubes and working fluid has been compromised. The cause could be fouling on the tube ID (often caused by mineral build-up), tube OD fouling, or a combination of the two.

 

[Asymptotic]

Has the pump been taken apart and inspected?
I've observed LR pump overheating due to the gradual build-up of solids contaminants between impeller and housing.

 

[Amused_1]

Ok, I’ll respond to your first response in the morning after I’ve refreshed myself on some process physics! I should probably mention that while I’m a Lead Operator, we actually have a Technical Manager that’s supposed to be sorting all of this. He’s distracted with another project at the moment so I’ve taken it into my own hands.

The pump is due to be inspected ASAP, but we would have to run the vacuum for the plant through the High Vac. This is entirely doable, but we’ve had a few production runs that require both vac systems to be used in order to meet customer specifications.