Forums
Blogs
Articles
File Library
FAQ
Search
FB
I have been looking at chiller suppliers, and the use of chilled water/glycol in vent condensers used to knock out VOCs.
Chillers themselves are rated to give a specific amount of cooling (kW) at specific inlet and outlet temps, for example chilled water returning to the chiller at 5 degC, but being sent back out to the plant at 0 degC. In this case a delta T of 5 degC seems reasonable.
Regarding vent condensers, heat exchanger suppliers, have proposed designs where the delta T of the chilled water is only 0.5 degC. This results in a very high flow rate of glycol.
I believe that they like to keep the flow rates high to increase the value of U, and have heard that chillers "like" to run on lower, rather than higher temperature differences. However, should I have specified that a 5 degC temp delta is allowed on the condenser - as the chiller is to run on 5degC delta T. Hope this makes sense.
|
|
Refirgeration
Started by Guest_fred_*, Apr 07 2004 06:41 AM
1 reply to this topic
Share this topic:
#2
Art Montemayor
-
- Admin
-
- 5,782 posts
Gold Member
Posted 07 April 2004 - 11:06 AM
Fred:
I find that your 5 oC delta T on your chilled water is rather low. I would expect (because I've done it on vent condensers) to supply 0 oC chilled water to a condenser and obtain 10 oC return water back to my chiller. That is twice as much as what you are asking - and I believe I could obtain more.
I believe it more reasonable to expect (& obtain) a 10 oC Delta T on the chilled water. I say this because heat transfer in a tubular heat exchanger (this is what I assume you're applying) is physically dependent on the temperature approach that the heat exchanger can deliver - not the temperature delta T itself on the chilled water. What I mean to say is that the delta T is a result of the feasible and obtainable temperature approach. Assuming an ideal, counter-current vent condenser (for the sake of discussion, only) the approach of the return chilled water would be the temperature of the return water proximate to the entering warm vent gases (which I would assume as ambient --- 25 oC). To expect that the return chilled water could get close within the inlet vent gases' as to be 15 oC apart is, I believe very reasonable and within practical achievement. The return chilled water temperature, then, is dependent on the temperature of the inlet vent gases and the efficiency of the heat exchanger you employ. Of course, if your vent gases are at 5 oC, you should fully expect to have a 1 oC return temperature on the chilled water. To give you an example, I have designed (& obtained) heat exchangers with an approach of 2 oF (1.2 oC). I certainly don't recommend you go to this extreme, but it can be done. All it means is a bigger and more expensive exchanger. But you don't have to have an expensive exchanger and yet obtain an approach of 10 oC.
You, as the operator, want to have as low a chilled water flow rate as possible (this translates to as efficient a heat exchanger as you can obtain for your money). You should specify to your supplier the temperature approach that you desire on the return chilled water. The flow rate of the chilled water has nothing to do with the Overall Heat Transfer Coefficient, "U". The "U" is dependent on the film coefficients of your system and the chilled water velocity - not the flow rate. I can obtain a higher "U" with a lower flowrate by simply increasing velocity (reducing cross-sectional area) -- but of course I would have to pay with more pressure drop.
Chillers don't "like" to run, period; but they will, when specified accordingly. From the heat transfer point of view, though, a chiller will transfer heat faster with a larger LMTD (Log mean temperature difference) - or, conventionally, delta T. Therefore, I don't understand why you state that chillers "like" to run on lower, rather than higher temperature differences. The basic heat transfer equation still is:
Q = U A (LMTD)
Both U and A are constant (or can be expected to be as such); that means that as the LMTD increases, so does the Q (heat transfer, in Btu/hr).
There may be other basic data that you haven't revealed that may be driving this problem in your direction. But I would normally expect that you would specify (& obtain) a 10+ oC temperature on your return chilled water. I hope I have been of help rather than confusing the issue.
Art Montemayor
Spring, TX
I find that your 5 oC delta T on your chilled water is rather low. I would expect (because I've done it on vent condensers) to supply 0 oC chilled water to a condenser and obtain 10 oC return water back to my chiller. That is twice as much as what you are asking - and I believe I could obtain more.
I believe it more reasonable to expect (& obtain) a 10 oC Delta T on the chilled water. I say this because heat transfer in a tubular heat exchanger (this is what I assume you're applying) is physically dependent on the temperature approach that the heat exchanger can deliver - not the temperature delta T itself on the chilled water. What I mean to say is that the delta T is a result of the feasible and obtainable temperature approach. Assuming an ideal, counter-current vent condenser (for the sake of discussion, only) the approach of the return chilled water would be the temperature of the return water proximate to the entering warm vent gases (which I would assume as ambient --- 25 oC). To expect that the return chilled water could get close within the inlet vent gases' as to be 15 oC apart is, I believe very reasonable and within practical achievement. The return chilled water temperature, then, is dependent on the temperature of the inlet vent gases and the efficiency of the heat exchanger you employ. Of course, if your vent gases are at 5 oC, you should fully expect to have a 1 oC return temperature on the chilled water. To give you an example, I have designed (& obtained) heat exchangers with an approach of 2 oF (1.2 oC). I certainly don't recommend you go to this extreme, but it can be done. All it means is a bigger and more expensive exchanger. But you don't have to have an expensive exchanger and yet obtain an approach of 10 oC.
You, as the operator, want to have as low a chilled water flow rate as possible (this translates to as efficient a heat exchanger as you can obtain for your money). You should specify to your supplier the temperature approach that you desire on the return chilled water. The flow rate of the chilled water has nothing to do with the Overall Heat Transfer Coefficient, "U". The "U" is dependent on the film coefficients of your system and the chilled water velocity - not the flow rate. I can obtain a higher "U" with a lower flowrate by simply increasing velocity (reducing cross-sectional area) -- but of course I would have to pay with more pressure drop.
Chillers don't "like" to run, period; but they will, when specified accordingly. From the heat transfer point of view, though, a chiller will transfer heat faster with a larger LMTD (Log mean temperature difference) - or, conventionally, delta T. Therefore, I don't understand why you state that chillers "like" to run on lower, rather than higher temperature differences. The basic heat transfer equation still is:
Q = U A (LMTD)
Both U and A are constant (or can be expected to be as such); that means that as the LMTD increases, so does the Q (heat transfer, in Btu/hr).
There may be other basic data that you haven't revealed that may be driving this problem in your direction. But I would normally expect that you would specify (& obtain) a 10+ oC temperature on your return chilled water. I hope I have been of help rather than confusing the issue.
Art Montemayor
Spring, TX
