4 replies to this topic

Dear All

Our client has asked us to investigate the feasibility of reducing the process water outlet temperature from their existing cooling towers by replacing some of the air inlet stream with available waste dry nitrogen. The plant heat load (of which we do not have complete details but can estimate) is primarily made up by their main 4 stage intercooled centrifugal compressor which rejects about 19MW. Hence the intention was to reduce the process water temperature, thus increasing the intercooling (due to an increased LMTD) and reducing the power consumption by this compressor (and other equipment) and realise a cost saving.

We have already calculated that they do not have nearly enough nitrogen to be able to achieve any savings. They have about 20,000Nm³/hr available and the air flow is approx. 4,000,000Nm³/hr hence this would be a drop in the ocean. I think I have a pretty good understanding of the thermodynamic interactions that occur in this type of system but there is one thing that I am not so sure about. Will adding nitrogen to the inlet air stream reduce the temperature of the process water oulet temperature or not? The logic behind believing this is as follows: by introducing nitrogen we decrease the humidity (and wet bulb temperature) of the incoming stream and hence it should be able to absorb more energy with all other conditions remaining the same. Presumably the process water outlet temperature will decrease to a point where equilibrium is reached based on the new humidity/wet bulb temperature. At equilibrium the range will be the same at some lower temperature? Alternatively depending on the cooling tower characteristics the air/nitrogen will just leave at a lower enthalpy and hence will not have absorbed any more heat than before adding nitrogen?

I understand that if we model this system as a steady state steady flow system then the cooling tower will only be able to remove the amount of energy that goes into the system i.e. the plant heat load is what determines how much heat the cooling towers can extract and hence the range. But as we introduce nitrogen we are no longer dealing with a steady state system as we have decreased the incoming energy on the air side and hence should be able to decrease the outgoing energy on the water side.

I have not designed cooling towers so do not fully understand the use of demand/performance curves (we do not have the performance curve from the coooling tower suppliers) but I hope this query should be able to be answered theoretically...please help?

[Art Montemayor]

Alistair:

The proposal is to add essentially bone dry air to partially humid air entering a water cooling tower and obtaining cooler outlet water. The amount of the added dry nitrogen is 0.5% of the total air volume entering.

In order to hopefully obtain a cooler water product, what is proposed is the wrong way to configure the operation. By mixing the nitrogen with the incoming air, the driving force of the humidity is getting diluted. A more effective way is to add a cooling tower in series and let the dry nitrogen enter the second tower which would be downstream of the first. That way, the bone-dry nitrogen is more effective and the equilibrium between the inlet nitrogen and the outlet cool water is maximized with a colder water being produced. But that takes a duplication of cooling tower investment – and you have to have sufficient nitrogen to make it worthwhile. That is the trade off – and an expensive one.

I would make a quick estimate of the heat pick up assuming the above: a heat balance of the second tower would have the heat picked up leaving with the saturated nitrogen equals the added refrigeration in the outlet water. This is the best of all theoretical worlds – and I’ll bet you wind up with a very small cooling effect and the ultimate result that it isn’t worth the effort. You should be able to come up with the enthalpies of saturated nitrogen and dry nitrogen. The enthalpies of the water are easy. Make sure your enthalpy bases are the same when making the balance.

I remember doing the same exercise on my air separation plants about 40 years ago.

Hi Art, thank you for your (as usual) concise and informative reply.

I have done the maths, all seems ok. The tables I have (Cengel & Boles. Thermodynamics An Engineering Approach 3rd Edition. McGraw Hill. Tables A-4 & A-18)seem to use a different base for water and nitrogen (0kJ/kg at 0^oC for saturated liquid water and 0 kJ/kmol at 0K for nitrogen) hence I have calculated the heat rejected using 3 different approaches with all yielding similar results (from 365 to 381kW). Surely, since I am dealing with a change in enthalpy then the base doesn't matter? I am only concerned with the change in energy in the nitrogen (or air) stream equaling the change in energy in the water stream?

I get about an additional 0.4MW rejected which equates to a 0.1^oC drop for adding another cooling tower with Nitrogen only; and an additional 0.24MW rejected when considering replacing some of the air with Nitrogen which equates to a 0.06^oC drop. When considering the power saving on the compressor (85kW) it is clear that this is most certainly not a viable option. If they had enough nitrogen to get at least a one degree drop it would probably become feasible.

It must be noted that these calculations are based on the instant at which the nitrogen is introduced (unsteady state) and purely on energy and mass balance calculations. So the actual power saved would be less than 85kW and the actual temperature drop would be less than 0.06^oC once the system reached equilibrium. To calculate the actual energy saved and temperature drop at equilibrium, one would need to know the overall heat transfer coefficent of the entire heat load as a function of temperature as well as to have the cooling tower performance curves; this would be a considerably challenging thing to do in reality. Does anyone agree/disagree/comment? Has anyone tried to do this on a complex system before?

Also, the Merkel equation on which most cooling tower designs are based, assumes a relative humidity of 100% on exit - how realistic is this? I presume it depends on many things. It's just that the installed system has a heat load of about 40MW but using the above assumption the cooling towers should have a capacity of over 60MW but they are still having problems with too high temperature process water.

Art, did you come to the same conclusion some 40 years ago? Not enough nitrogen?

Thanks again.

[Art Montemayor]

Alistair:

Yes, that was one of the pronounced conclusions that I remember from my attempt 40 years ago. The other conclusion was that I needed more tower capacity. These two negative prospects allowed me to "can" the proposal immediately after doing the math.

My scope was to improve the Liquid Oxygen production performance and I did this by concentrating on the compression inter and aftercoolers on our main compressors. I built spiral tubular exchangers that reduced the air outlet temperature 3 oF from the original, and I got what I was afteer.

[HGalaviz]

Question?

Why not have a cooling tower company come out for a site visit to look at the cooling tower? These days the cooling towers are like cell phone, they are getting updated everyday. There is so many improved products now for better cooling all you need to do is find a good company that cares and ask. If you would like for me to recommend you one just let me know.