Cooling Tower Evaporation Question.

I'm sure you've already considered this, but if your tank was exposed to the severe cold, could it have damaged your high level alarm?

The lower levels would have been prevented from freezing by the circulation, but if condensation froze solid in the upper level......

NO. the tank is inside. (we will be putting in different alarms in the tank also, they are discreet now, but we will be upgrade them to radar later).

And the controls show they are working.

On a added note, when I was first told this, at first I thought maybe the make-up water valve was stuck open, even when the control is off. But that doesn't look like the case.

Your thoughts seem to be valid and correct.

The make up water (from seals) is probably greater than the evaporation. The cooling cells (67 to 60deg) are probably running extremely efficiently given the ambient air temperature such that no evaporation is actually happening.

Solution suggested is a simple overflow relief set 1/4" above the high level alarm signal. This "fail to safe" situation will then provide signal if the make up valve is left open, but will ensure that excess water is directed to where you want it to go.

Saw similar in injection moulding shop heat exchanger during winters over here. Return water was colder than required and we actually had to use local controllers at each machine to "shandy" the tower water with the process water to achieve stable temps inside the tools.

The make up water (from seals) is probably greater than the evaporation.

The problem I have with this, is that the cooling effect is from evaporation, unless now with the cold outside temperature out side some cooling is now from conduction. That I would understand and accept it.

About the overflow relief........We had an open port at the top of the tank,........

.... had to use local controllers at each machine to "shandy" the tower water with the process water to achieve stable temps inside the tools.....

I have a throttling valve on the main to do that. but for us, cooler the better. But we spec'd it using 75deg F wet bulb.

on the up grade, I'm am happy that we put in the instruments or at least ports to grab data off each process to trouble shoot....... The new process isn't even on-line yet, but we already started to use these on the process.

Thanks for your input,

If your entering dry bulb temp is below the water temperature, sensible heat transfer (not sure whether this would be considered conduction or convection) will take place, replacing some of the 'normal' latent heat transfer. The higher the difference in temperature, the higher the sensible heat. But there will always be at least a little evaporation, as long as the exiting air temp is higher than the entering air temp, even if the entering air is 100%RH.

Yes there is evaporation...... An a real cold you can easy see it...... It look like a snow making machine. Frosted up quite a bit. I just don't know if 80 GPH would do this.

Just for clarification, we were using cooling towers to cool liquid (water/coolant mix), not air.

Yes

A trip down memory lane for me.

I was involved in sales and service of (fairly small) evaporative and air-blast water coolers, initially for use with water-cooled compressors and after-coolers, and much later followed by water-cooled fridge-dryers.

Industry caught on to the fact that 'cheap' water-coled devices, brought in their wake, relative costly-water cooling systems, especially when water companies found they could charge premium rates for supplying the original water.

Eventually direct air-blast coolers became the norm, thus avoiding water problems completely.

My overall experience in after-sales nearly always related to lack of cooling blamed on the coolers being 'too small'. There was some merit in that logic because the initial cost led to saving money by buying the 'minimum' cooling necessary. Any increase in the 'heat load' (usually due to ambient conditions) soon caused over-heating. Applying the 'real' ambient conditions to the original cooler selection soon revealed where the problem was.

Whatever the reason for lack of cooling in past cases, I ask here if your cooling system is working in terms of providing cooing to your equipment. Good! because that is the prime function.

The focus will be on were the water is going (or not going) to cause the high level alarm. Which seems to be the case

Well, we added a new cooling tower, but we had to take it off line, because the trail pitch was set too steep.

We don't have the new process on line, so out the cooling requirement as are the same.

The only change was the amount of supply to the process, (Larger pumps, more Volume, I calculated the system pressure and the requirements was about 2/3s of the original. Other then that nothing changed. So the questions was, why were we getting the high alarms...... (btw, We turned down the seal water requirements, we have flow control on these and some were peg wide open.) But I shut off the make up water and it took 16 hours to bring it back in range, (between the high level and low level sensors which is about 8" apart)

As I mentioned earlier, we will be going with a radar level sensor, when the next phase is in place. The capability will be 9,000 GPM. with an extra towers. (But currently, we have (3) cooling cells 1500 GPM, 1500 GPM and the new one that is locked out until the trial blade pitch is corrected is spec'd at 1850 GPM)

To phoenix.

When I said memory lane, it was, the 1970's in typical UK ambient conditions. Quite honestly the size of the products I was involved in were very small and terms you use like 'pump seals' and 'trail pitch' never cropped up in the work I was doing. So I can't help here if this is the area you suspect.

But with my limited knowledge I checked water pressure and temperature differentials at the inlet and outlet of each heat exchanger and matched this to the water flow rates to see if the heat balance was correct. An unbalance dropped hints where to look.

In one memorable incident every calculation pointed to lack of water flow (which defied logic) but was proved true when eventually the flow control valve was taken out and inspected. A rubber glove was found tangled up in the mechanism.

The fitter in charge of cooling tower maintenance admitted he has lost a glove and was pleased to get it back - until he was told where it was found. Whoops!

As I said, memories. and perhaps outdated compared to modern practice, but I guess fundamental engineering principles are much the same today.

Compared to my compressed air work - water was a doddle to deal with - you could see it!

Differential expansion and contraction? Consider the possibility of the piping and the tank getting smaller, with applied cold, than the contents, thereby increasing the tank level. Increased system volume through the new process will exacerbate the problem. The solution is to lower the maximum operating fill level set-point so that the alarm level is not achieved when the lower temperatures are experienced. Why not make the low level alarm into a normal operating point (rhetorical question - NNTR)?

• If it were a domestic central heating header tank this mission would be as simple as bending the ballcock's Portsmouth Valve's operating lever downwards so that it shuts off at a lower level. Ballcock Portsmouth Valve operating levers are made of bendy metal specifically for this purpose.

If the operating level set-point is already at the minimum possible, then a larger tank with a similar low set-point is the solution.

That's why, I mention earlier, we are switching out the discreet level controls for radar, we would have more control on the levels for a variet of reasons.

Could it be that the sensors are located in a turbulent zone and water surface waves are triggering the alarm.

The radar sensor might need some baffles and such to provide a smooth surface for sensing. (Similar problems in our water treatment plants and reservoirs.)

No, we also have a visual sight. (Clear hose that runs on the outside the tank). Our control engineers have things under control..... But you are correct, with the turbulent flows, this is the reason we like to adjust the level, the pumps are running about 45%. We would like to be able to raise the level. To settle out any air entrainment in the cooling water.

"When low level is reached pump make-up water is added and is not shut off when high level alarm is reached. (There is a third sensor below the low level alarm if for some reason water level gets too low)"

What shuts it off? What provides a signal for the make up pump to shut down?

We just changed the configuration temporary, with out going into the reasoning. We have three sensors.

- A high level alarm, which shuts off the makeup valve.

(Alarm may not be the best terminology, but it's the level control sensor)

- a low level alarm which opens the make up valve for filling.

- and a warning alarm below low level alarm. Just in case for any reason, it reaches this, where a we lose water due to any seal failure, or even make-up valve failure where it would shut the process system down.

The problem we were having, is that we were getting a high level alarm, without the make-up valve opening.

Even though there is a warning light that indicate there is a problem. I just had to find, where is the supply coming from that's doing this.

The only outside source (other than a possible HX failure in the process, but that is ruled out) is the pump seal water.

Is there the possibility that the 85GPM fresh water used at the pump seals is greater then the system losses?

Would also look at how the return from the cooling cells enter the tank. Turbulence do to the increased flow could be setting it off.

Yes, that what I believe is happening. the seal water make up is greater than system losses.

As far a the turbulent flow in the tank.

That is one of the reasons why we are putting in radar controls, to vary the tank level. But the problem is air entrainment, we want to settle out the water by changing the level. The pumps itself only require about 6ft of suction head, (Which is unbelievable for what they can do.)

But the reason why its not that, is that one time, it actually filled up the tank and went into the over flow.

Being not knowledgeable of how the tank was constructed. On tanks I've dealt with mostly hydraulic. The return line is pipe to the bottom internally. This aids in keeping the oil from foaming.This may aid in the air entrainment if the return water from the cooling cells is just being dumped in the top. Would also reduce turbulence at the surface.

The answer is elementary, Dr. Watson. When the outside air is -20 F, there should be little to no evaporation demand placed on the water, in fact, when we get to temperatures below freezing in our power plants, we have to spin the cooling fans backward to hold heat in the cooling towers. So, I expect that your makeup water demand is very near or below the seal water (that goes to makeup), and you made no mention of blow-down on the cooling tower. Perhaps you need to install a blow-down controller, and have the system blow-down based on conductivity and on time, to keep some throughput in all conditions.

Thanks,

I didn't mention blow-down, unless someone mentioned it, as to not overwhelm or polluted my description of the problem. Yes we have one that is connected to a controller already.

Thanks for mentioning it.

We made adjustments on on the supply of our seal water. We will be bringing another process on line, of which will put more load on the system. But that happens we will be increasing the GPM.

(Also we have one cooling cell with the fan off, Its a new one w just installed, but the trial angle was set too steep and our supplier has to correct that, motor was drawing too many amps). and at times we will shut the fan off on a second cell, leaving one cell running.

But we have to be careful with that, due to ice build up.

Thanks for your input.

thwill be tuthat is on line

Let's take another look at the seal water:

What are the actual quality requirements for the seal water? Could you not just filter the basin water extremely well and/or soften it, and eliminate the additional flow into the system? Does the seal water have to go into the system at all? Could it be on a closed loop cooler? Closed loops are easier to treat and maintain in a non-corrosive, non-scaling state.

This problem only exposed an opportunity to improve conditions.

I'm believe this is only a temporary situation.

My initial concern was if there was a leak in one of our HX's in the process that was added to the cooling water. But I've eliminated that.

I did not realize that the seal water would over power the evaporation losses and I just wanted to confirm that.

But I did go through and made adjustments to the seal water flow to some of the pumps, some were opened too much, and since then we haven't had a problem.

Cooling waters treated of course by our water rep. and we like to keep it simple, adding to the complexity and not gaining anything we want to avoid.

I believe we have it in hand now, Anyways here are the seal water requirement before the adjustments..

If quality of the seal water is not an issue, why not use the cooling tower water for seal water? OK, maybe you would need to filter that, but why not use it?

The quality of the seal water is important. Don't believe I said it wasn't.

The reason was, if we were needing make-up water any ways, use the seal water. The source is the same as the make-up water.

It's just that with the up grade, we double the GPM of the cooling water, thus getting better efficiency out of the existing system.

When the new process comes on line, and we are still having problems we have other alternatives we can use.

One of what, we can control our tank level giving us more reserve space, as I describe earlier the new pumps require less suction head.

Among other things, and if this is an issue, with the little amount that does run over, it would go out on the tank over flow......... its just that with the weather, our over flow to the outside is froze up.

There are many alternatives, again my biggest concern was, could the seal water over take the make-up requirements. And it could and did.

And again, I adjusted the seal water on the pumps (pumps 4010 and 4160 from chart) that dropped at minimum 10 GPH as make-up water, and we have had -0 temps, never had the problem so far, and it's been a week.

So right now, I'm putting it as problem solved and no longer an issue, because of the value of the responses, I just followed up on the responses as a common courtesy,