PDA

View Full Version : Esceeding chiller flow specs


ReeferAl
06/22/2006, 04:25 PM
I'm getting ready to hook up my Aqualogic Deltastar chiller. It's a 3/4hp with flow rating of 900-1800gph. Is there a reason I can't have more flow than that through the chiller? I previously had it hooked up to an AM3000 pump with about 2500 gph of flow with no apparent problems. The new pump will put out about 50% more flow however. I would expect it would be about twice the flow recommendation. Has anyone done this and if so, any problems?

Allen

Fiziksgeek
06/22/2006, 05:39 PM
I dont ave any experience with this, but I do have some insite. Each ciller as a range of flow rate that allow te chiller to perform ats peak efficiency. If you are outside of this range, you are simply decreasing its efficiency. The water will be flowing too fast, and won't give up as much heat.

If you have a real heat problem, your chiller may not take care of it. Your 3/4p ciller may be acting like a 1/4hp....

xtrstangx
06/22/2006, 05:40 PM
You could put more flow through it but you wouldn't be using your chiller's optimum performance. Your chiller would likely have to run longer and you'd be wasting electricity doing so.

wlagarde
06/22/2006, 11:24 PM
Reefer AL - Remember your Med-School lessons on counter current exchange....the skimmer will perform fine with excess flow. In fact it will perform more efficiently than with less flow. The tubing size of the exchange element generally limits the maximum flow. Trying to deliver flow beyond the max rating results in exponential increases in back pressure...

Fiziksgeek
06/23/2006, 06:05 AM
<a href=showthread.php?s=&postid=7612292#post7612292 target=_blank>Originally posted</a> by wlagarde
...In fact it will perform more efficiently than with less flow. The tubing size of the exchange element generally limits the maximum flow. Trying to deliver flow beyond the max rating results in exponential increases in back pressure...

There is a very valid point here. I did not think about the size of the tubing in the heat exchanger. Its kept small to increase surface area.

However, there is more to it then that. The amount of heat extracted is dependent upon the delta T (temperature difference between the hot thing and the cold thing) and the contact time (how long they are touching). There are of course other factors, like what are the materials, but they are pretty much fixed, salt water and titanium. Long story short, you absolutely will be decreasing efficiency by pushing more water through it. That of, course, doesn't mean the chiller won't meet your needs.

What is the new pump you are planning on using? You said you had been running it with about 2500gph, and the new pump was about 50% more...meaning about 3750gph? Sound like it might be a Sequence Dart? Maybe an AmpMaster 3000? With those pumps, your will have so much back pressure it will keep the flow within the chillers spec (just guessing here).

Reefski's
06/23/2006, 06:08 AM
it may work but a big waste of electricity with an oversize pump.

wlagarde
06/23/2006, 07:34 AM
carloski - YOU ARE COMPLETELY WRONG. Moving water through the chiller faster INCREASES the delta T between the aquarium water and the chiller element so therefore it INCREASES the efficiency...

Beenalongtime79
06/23/2006, 07:35 AM
I agree, by using such a large pump on a chiller that isn't designed to handle that flow, you are probably creating much backpressure on the pump you will be using which will basically thwart the idea as you will get maybe slightly more flow at an increased cost. Why don't you just T it off... that might make your life much easier.

Peace,
John H.

Charlie Davidson
06/23/2006, 07:47 AM
Moving water TOO Fast--- decreases efficiency.. The water is moving thru unit so fast it doesn't have time to remove heat. (been working on chillers @ 30yrs)

wlagarde
06/23/2006, 08:05 AM
Charlie - Again wrong. The water is still in thermal contact with the exchange element so heat transfer occurs without a problem. What happens is as follows: Lets say the water flowing through the chiller is 500gph and you measure the temperature of the inlet and outlet water and find a delta of say 1 degree. If you double the flow the delta of the temperature of the inlet and outlet water would be at least 0.5 degrees. 500 x 1 = 500; 1000 x 0.5 = 500 same amount of heat removed. However, the delta T between the inlet and outlet with increased flow under the conditions outlined above would likely be greater than 0.5 degrees...say 1000 x 0.55 = 550 thus yielding greater efficiency. As flow increases the amount of heat extracted will increase and asymptotically reach a maximum. Again, increasing flow INCREASES the delta T between the exchange element and the aquarium water thus improving efficiency. This is actually simple physics. Take a look at this drawing: http://en.wikipedia.org/wiki/Countercurrent_exchange

Charlie Davidson
06/23/2006, 08:09 AM
only till you reach a certian point! (been there done that, seriously)

wlagarde
06/23/2006, 08:15 AM
Exactly - That point is a maximum that is approached asymptotically...the gain would be minimal. I only make the point to say increasing flow will not compromise efficiency of the chiller. So, you can try to bring more flow though the chiller. However, if you are plumbing through the chiller in the return of water flow to the tank return flow not efficiency will be compromised...

Fiziksgeek
06/23/2006, 08:37 AM
Physicist here....Like I said before, heat exchange is not only a function of the delta T, but the contact time. You will indeed increase the delta T by pushing more water, but you are decreasing the contact time.

You absolutely will get to a point where the benefit of increasing the delta T will be outweighed by harm done by decreasing the contact time.

Beenalongtime79
06/23/2006, 08:39 AM
Doctors can be such wankers. Haha

wlagarde
06/23/2006, 08:50 AM
Nope - Water is in constant contact with exchanger - Physician/Engineer here...

ReeferAl
06/23/2006, 09:21 AM
I agree with wlagarde. With less flow you will have less water bing cooled more. With more flow you will have more water bing cooled less. The result will be more or less the same when the water is then mixed with the rest of the system water.

As far as the pump being used, I was previously using an AM 3000. Actually I had 2, 1 going directly to the tank and the other through the chiller. Water is also teed off to the grow out tank and refugium. The flow I gave before was an estimate based on head loss estimates.

The new pump is a sequence barracuda. I hooked it up and throttled it back some. Even so the flow into the tank increased considerably. I then turned off the other AM pump and opened up the barracuda valve a bit more (but not all the way). The flow was still greater than it had been with 2 AM 3000 pumps running. It appears that the AM pumps are not putting out as much as they claim. Maybe it is because they are a few years old (although recently cleaned). Maybe they drop off more with the head loss.

In any case the chiller is NOT impeding flow to any significant degree. It appears to me that the heat exchange occurs by flow through a large pvc tube (4"?) which contains a titanium coil chiller element. I don't see any place where the chiller would actually impede flow to any degree.

Allen

Beenalongtime79
06/23/2006, 09:23 AM
Ahhh, so basically it is chiller barrel unit. Nice.

futrtrubl
06/23/2006, 09:30 AM
Yup efficiency will not be decreased (within the chiller) by increased flow (deltaT increased, contact time per molecule is irelevant, contact time with the water is still the same ie constant). However, if you look at the energy used by the pump then after a certain size the efficiency increase in the chiller will be outweighed by the extra energy used by the pump (and perhaps heating the water).
So while there are no pressing reasons to not use a large pump, there are no reason to use a large pump (unless you want this as your return or something).

Edward

Fiziksgeek
06/23/2006, 09:37 AM
<a href=showthread.php?s=&postid=7613693#post7613693 target=_blank>Originally posted</a> by wlagarde
Nope - Water is in constant contact with exchanger - Physician/Engineer here...

So why can I quickly touch a red hot burner on my stove and not get burned?

For a small amount of time, my finger is in intimate contact with the very very hot burner, huge delta T! And, as a physician would know, our bodies are mostly water, salty water.

Heat exchange takes time.

And if the exchanger is basically a straight tube with the coil in it, thats even worse. You flow will be pretty laminar...only a small portion of the water will actually contact the cool surfaces, most of the water goes in and out, and is not cooled at all.

wlagarde
06/23/2006, 09:43 AM
Yes but Fiziksgeek - You must realize the contact time of the water with the heat exchanger is constant (100% of time) regardless of the flow. Flow increases the delta T. If you increase the flow of water from 0 to infinite over a fixed period of time, contact time does not change. What changes is the delta T and thus heat transfer increases until you reach a point where the temperature of the exchanger equals that of the aquarium water. THis is the point of maximal efficiency. The chiller's max capacity then is limited by the flow of coolant through the exchanger...

Lo0seR
06/23/2006, 10:02 AM
This thread is funny, I used whatever pump was available at the time for my 1/4hp chiller which was a mag18 last week. It got to 85 water temp in my prop tank in the garage (200+gal), it took forever to bring the water temp down, ya I know 1/4hp is a little to small for that tank but it works. Here is the kicker, went to the LFS store the next day got a quietone 4000 because the mag18 was just to much flow and blowing stuff all over and the specs said 7GPM anyways for the chiller. Switched out the pumps, flow is now up to specs, first thing I noticed is chiller hardly comes on now, not blow stuff around and water temp stays a nice 79 to 80.
So I think low flow is the way to go.

Reefski's
06/23/2006, 10:05 AM
simple law of physics. you can only push so much water through a given pipe size. backpressure = wasted electircity which will shorten the life of your pump and waste your $$$ on electricity.

i bet the engineers who designed the chiller have actually done the calculations to tell you the max and min size pump to use so you get good efficiency without wasting your money on an oversize pump.

wlagarde
06/23/2006, 10:11 AM
Agreed on this point...however, one could simply do the experiment...increase the flow and measure the backpressure...if it's not too high then you are OK.

Fiziksgeek
06/23/2006, 10:28 AM
HAHA, I actually enjoy conversation like this. Sometimes I'm wrong, sometimes I'm not. In this case, I am pretty confident I am correct :-)

I hope I say this right.

Let's assume the coil is an infinite heat sink, meaning it always maintains its temperature (cold). When a molecule of water comes in contact with the coil (titanium I assume), it begins to transfer heat. It can only give up heat until it is at te same temp as the coil itself. Lets say this water molecule is in contact for 1 second, and give up 2 degrees of temperature. We push the water twice as fast. So contact time drops to 1/2 of a second, you migt only give up 1.4 degrees of temperature. So, in order to get that full 2 degree of temperature drop, the water has to go around 2 or 3 or 4 times. Dont forget, the delta T does have a significant effect, so the closer you get to the coil temperature, the smaller the delta T is, and the heat you extract. So, the second time around, at high flow, will not produce the same temperature drop that it did the first time.

Obviously, I have just made up numbers here, so no one do any calculations with these specific values and come back and tell me I am wrong.

Lo0seR--that is why wen you reduced flow, your chiller worked less. The water gave up more heat on each pass, no need to send it around and around.

wlagarde
06/23/2006, 10:37 AM
Again - wrong - I sugest you do this experiment: Hook up a flow meter, gate valve, and pump in line with you chiller. Take measurements of the temperature of the water entering the chiller and leaving the chiller at say 100, 200, 300, 400, 500gal/hr, etc. Calculate the delta T between the input and output water and multiply this product by the respective flow rate for each measurement x 3.785kca/gal to calculate the heat removed by the chiller in kcal/hr. You will see the amount of heat removed increases as flow increases until a maximum is reached. At this point with further increases in flow the heat removed in kcal/hr will not change either up or down...

Fiziksgeek
06/23/2006, 10:44 AM
<a href=showthread.php?s=&postid=7614158#post7614158 target=_blank>Originally posted</a> by carloskoi
simple law of physics. you can only push so much water through a given pipe size. backpressure = wasted electircity which will shorten the life of your pump and waste your $$$ on electricity.

i bet the engineers who designed the chiller have actually done the calculations to tell you the max and min size pump to use so you get good efficiency without wasting your money on an oversize pump.

But if you could increase flow to make the chiller more efficient, I would rather run a pump that draws 2 amps longer, and reduce the time the 13amp chiller was on. (of course there is a alays a break even point).

My argument is that you are not increasing the overall efficiency of the chiller by increasing the flow. So not only are you using a pump that uses extra power, your chiller will probably be on longer than is really necessary.

Fiziksgeek
06/23/2006, 10:50 AM
<a href=showthread.php?s=&postid=7614354#post7614354 target=_blank>Originally posted</a> by wlagarde
Again - wrong - I sugest you do this experiment:

Not wrong...done similar experiments, not on aquariums, but on lasers. Same situation, got heat and want to get rid of it...

Again, if it was as simple as increasing flow, why wouldnt the manufacturer tell you that?? It would be a huge selling point!

wlagarde
06/23/2006, 11:52 AM
yes wrong - Actually performed experiments with a chiller in a laboratory.

futrtrubl
06/23/2006, 12:15 PM
<a href=showthread.php?s=&postid=7614299#post7614299 target=_blank>Originally posted</a> by Fiziksgeek
Let's assume the coil is an infinite heat sink, meaning it always maintains its temperature (cold). When a molecule of water comes in contact with the coil (titanium I assume), it begins to transfer heat. It can only give up heat until it is at te same temp as the coil itself. Lets say this water molecule is in contact for 1 second, and give up 2 degrees of temperature. We push the water twice as fast. So contact time drops to 1/2 of a second, you migt only give up 1.4 degrees of temperature. So, in order to get that full 2 degree of temperature drop, the water has to go around 2 or 3 or 4 times. Dont forget, the delta T does have a significant effect, so the closer you get to the coil temperature, the smaller the delta T is, and the heat you extract. So, the second time around, at high flow, will not produce the same temperature drop that it did the first time.


Not exactly. Remember that the average deltaT for the pass is based on the average T of the molecule, roughly the average of the entry and exit Ts of the molecule. For slow flows this will cause the average deltaT to be smaller. Also, and more imprtantly, yes your molecule will have half the contact time (if doing double the flow) but you will have DOUBLE the molecules flowing passed per unit time, which cancles the halved contact time. This with the increased deltaT create more efficiency with greater flow.
This is ignoring any heat effects of the pump, which I stated in a previous post.

Edward

wlagarde
06/23/2006, 12:27 PM
"Not exactly. Remember that the average deltaT for the pass is based on the average T of the molecule, roughly the average of the entry and exit Ts of the molecule. For slow flows this will cause the average deltaT to be smaller. Also, and more imprtantly, yes your molecule will have half the contact time (if doing double the flow) but you will have DOUBLE the molecules flowing passed per unit time, which cancels the halved contact time. This with the increased deltaT create more efficiency with greater flow.
This is ignoring any heat effects of the pump, which I stated in a previous post"

This is an excellent explanation - Very nicely stated.

PrangeWay
06/23/2006, 01:04 PM
The first law of therodynamics indicates that the sum of the inital mass flow * initial enthalpy is equal to the sum of the exiting mass flow * exiting enthalpy. When broken down it looks as though as you increase mass flow, you will increase cooling to infinity.

BUT. As you increase the mass flow of the water, the mass flow of the coolant is constant, meaning the exiting enthalpy of the coolant is increasing. The refrigeration cycle in a chiller than lowers this coolant enthalpy back to it's initial value (ideally). There is a point, where the refrigeration cycle of chiller (with it's fixed input work, ie electricty) can no longer lower the enthalpy to it's initial value. So the h(i) and h(e) keep climbing the unit becomes less and less efficent, removing less and less heat, till it reaches a point where it is no longer functioning, the compessor is just constantly running for kicks.

The previous arguments main miss was in only looking at the heat exchanger part of the chiller.


PW

RichConley
06/23/2006, 02:01 PM
<a href=showthread.php?s=&postid=7614134#post7614134 target=_blank>Originally posted</a> by Lo0seR
This thread is funny, I used whatever pump was available at the time for my 1/4hp chiller which was a mag18 last week. It got to 85 water temp in my prop tank in the garage (200+gal), it took forever to bring the water temp down, ya I know 1/4hp is a little to small for that tank but it works. Here is the kicker, went to the LFS store the next day got a quietone 4000 because the mag18 was just to much flow and blowing stuff all over and the specs said 7GPM anyways for the chiller. Switched out the pumps, flow is now up to specs, first thing I noticed is chiller hardly comes on now, not blow stuff around and water temp stays a nice 79 to 80.
So I think low flow is the way to go.

The mag18 pulls roughly 140w. The Quietone pulls 50w. Someone else can do that math, but I'd bet that the extra 100w of constant heating going on in your tank had more of an effect than the flow.


Prangeway, the situation you suggest is only really valid in a situation where the heat load is greater than the capacity of the chiller.

PrangeWay
06/23/2006, 02:35 PM
Prangeway, the situation you suggest is only really valid in a situation where the heat load is greater than the capacity of the chiller.

Exactly! Because the coolants mass flow is fixed, you can only increase the enthalpy. The 'Heat Load' is what the refrigeration cycle can reduce from the Joules of Heat that has been removed from the water. By increasing the water's mass flow, you increase the exit enthalpy of the coolant (aka it's pulling more Joules out of the water) hence increasing the heat load. So there is an upper limit to the mass flow of water where any futher increases result in degrading the performance of the chiller.

Uh to put it even simpler, the more gph through a chiller, the more cooling you get, the more cooling the coolant needs, the more heat load you've put on the chiller, and there will be a max. A chiller doesn't just magically get rid of unlimited heat.

Theoretically this upper limit, of mass flow of the water ,before you degrade performance is what the manufacturer would list as recommended flow.


PW

wlagarde
06/23/2006, 02:43 PM
We are talking about the ability of the chiller to remove heat. Not the heat a larger pump adds if it is submersed in the water. If an external pump is used as a return and the chiller is plumbed in line, it adds no heat to the water AND the skimmer still removes the same amount of heat in spite of it receiving excess flow...however, it will will run at highr back pressure. Do the experiment I suggested.

RichConley
06/23/2006, 02:52 PM
<a href=showthread.php?s=&postid=7616081#post7616081 target=_blank>Originally posted</a> by wlagarde
We are talking about the ability of the chiller to remove heat. Not the heat a larger pump adds if it is submersed in the water. If an external pump is used as a return and the chiller is plumbed in line, it adds no heat to the water AND the skimmer still removes the same amount of heat in spite of it receiving excess flow...however, it will will run at highr back pressure. Do the experiment I suggested.

My point about the pump was that his experience could be easier explained by the change in pump heat than the chiller flow. (nothign to do with theoretical discussion)

Carry on.

RichConley
06/23/2006, 02:55 PM
Prangway, it seems to me (and this is in no way my expertise) that your argument is based totally on the tank being a constant temperature source.

it seems to me that if the chiller pulls heat faster wiht a faster pump, then the tank temp would drop faster, and the chiller would shut off. It should, theoretically, need to run much less often with a faster flow.

wlagarde
06/23/2006, 03:06 PM
gotcha...

PrangeWay
06/23/2006, 03:26 PM
Prangway, it seems to me (and this is in no way my expertise) that your argument is based totally on the tank being a constant temperature source.

This is true. Without the coolant type, flow rate & initial coolant temperature, I can't do the calculations to see if a xxx gallon tank cools xx degrees before the chiller reaches it's limit. However it is still imporant to realize that any gain is not free, thermodynamics is all about equilibrium, any quicker cooling gained by increasing your flow will be offset by higher coolant temperaure, increased compressor work, and reduced efficency.

If assumptions have to be made, I'd say the bigger the system, the more likely it is you should stick within the recommended flow of a chiller. I personally would stay in the recommended ranges if only becuase it was designated by a mechanical engineer who knows far more about thermodynamics than I do, and about the most efficent operating parameters of the unit he designed.

Beenalongtime79
06/23/2006, 05:03 PM
Ah ha, so those manufacturer numbers aren't just made up.

Thanks PW for giving another perspective on the issue.

In the end it seems like, if you're already trying to chill a system with a heat load to high for your chiller to begin with, that you should try to maintain flow in the suggested range, but if you've got a massive chiller that has way more cooling power than you need for your system, that you are pretty free to just have at it in terms of flow (within reason that is...)

I think that will help out a lot of people who read this thread and think that if I keep increasing the flow, then the chiller (no matter how small) will work for me to infinitum. Hehe. Obviously that is not the case as you have pointed out with your eloquent dissertation on the thermodynamics of our universe. Makes a lot of sense to me.

Peace,
John H.

Beenalongtime79
06/23/2006, 05:04 PM
wlagarde,

Are you a resident or attending at Chapel Hill? I'm thinking of going there for my residency eventually. Hehe

Peace,
John H.

wlagarde
06/23/2006, 05:13 PM
Actually faculty - but did residency and fellowship here. It is a great place (as is Charlottesville) - what are you specializing in?

Also, I'm not suggesting increasing a chillers flow wil compensate for an undersized chiller. However, increasing flow to an appropriate sized chiller will not compromise its ability to cool.

Beenalongtime79
06/23/2006, 05:24 PM
I'm really debating between anesthiology with possible fellowship in critical care or thinking about urology. I like working with my hands a fair bit so I'm still debating over it although I've got to decide soon since I've just started my 4th year.

What did you specialize in?

wlagarde
06/23/2006, 05:32 PM
Gotcha - UNC has excellent medicine, anesthesia, and urology programs. My only adice would be to apply and interview broadly so you get perspective in programs...

Beenalongtime79
06/23/2006, 06:27 PM
Thanks, I will do that. :)