Chiller - retrun water temperatures in a 2 port system 1

[remp]

I have 2 chillers working nicely on duty/stand by set up. Each chiller is rated for 2400kw and approx 100L/s each and CHW F&R temp of 6/12 Deg C. It is a primary /secondary set up so the flow thro the chillers is constant and the variable speed secondary pumps vary the flow to the building. Only one chiller ever runs at a time. The flow thro the chillers are fix thro an automatic balance valve (set and cannot be changed).

The building AHU's are a 2-port set up.

Here'e the problems, well so says an energy consultant we engaged recently;
When the chiller is running at low load we never get the chilled water return temperature back at 12deg C. The consultant reckons becasue it is a 2 port system the desing flow and return should always be 6/12 deg C.
It is usually about 8 or 9 deg C with only a 3 deg split and not the 6 deg split you would expect.
The consultant reckons there must be a by pass at an end of line loop in the building open that is causing this problem and letting CHW at 6 deg C mix with the return water form the AHu;s and reducing the overall CHW retrun temp.

My question is, if the building is at say 1/2 load and every AHU is only dropping the air by 50% of its design capacity does this automatically mean we have 50% the design water flow rate thro it....isnt the consultant automatically assuming the heating coil which is a heat exchanger has a perfect corelation between load and flow rate (direct proportion)....in other words half the flow and you half the capacity, 1/4 the flow and you 1/4 the capacity...is this true??? Isnt every heat exchanger different....Is there really a problem with the system set up?

 

[walkes]

You might get 12C back on the secondary, that mixes with the 6C primary temp.

In terms of the load balance at 50%, the flow and delta T should match your load overall. (So if you have 50% load, your delta T is 50% of expected then your flow is 100% not taking into account heat exchange issues) There may be some heat transfer issues in the coils at lower flow rates or lower entering air temperatures.

 

[remp]

The water back on the secondary is always 8 to 9 deg C....

 

[SAK9]

Looks like the well documented low delta T syndrome.Your assumption that flow goes down by 50% when load goes down by half is not correct as cooling coil capacity vs flow characteristic is non linear.Flow requirement may remain high under part load conditions due to a number of reasons such as:
1.below design set points
2.Oversized cooling coils and two way valves
3.Non uniform design delta across various AHUs/FCUs
4.Dirty cooling coils
4.Low heat transfer rate under reduced air flow conditions(coil face velocity of less than 1.5 m/s).

There is no easy solution to this problem.A very detailed investigation involving measurement of flows ,temperature rise across all ahus under partial load condition is required to establish the real reason.

 

[cry22]

Much ink ran on this topic, try to google low delta T syndrome and you will find a slew of articles.
here is one such article

http://www.taylor-engineering.com/d...posium AC-02-6-1 Degrading Delta-T-Taylor.pdf

 

[remp]

Just as I thought cry22 & SAK9....thanks very much

 

[Drazen]

over what sack mentioned, weren't you stated yourself that primary has constant flow while secondary is in variable flow regime?

it should mostly mean that you cannot have any direct relation between flow and delta t.

actually, with low delta t syndrom you are lucky if your chillers do not cycling on and off too much, and over the fact that such state cause increased maintenance trouble and deterioration, you can also have temperature instability, and when some lower capacity consumer are the only ones with demand, they might not be supplied properly.

in general, energy efficiency is not primary problem, though energy consultants will not accept it, as they push their aspect to be the only one to consider.

"having significant problem" mostly means that you need additional investment to rectify it, either installation of additional chiller with lower capacity, or cooling bank in form of chilled water tank, anyhow serious engineering solution would be required, but, as stated, investment would mostly be imposed by function issues, not energy efficiency because payback in energy analysis would likely be overkill.

 

[SAK9]

Remp,

It would be worth while checking Delta T across some of the large AHUs under full load and partial load conditions.In theorey Delta T under partial load has to be higher than the design value due to the non linear characteristic of the coil.But this seldom happens as flow control under partial load is not great due to lack of "authority" of the control valve.Retrofitting with pressure independent Control valves(PICV) alleviates the problem to some extent.See the attached for more info.In my opinion manual balancing valves should not be used on variable flow systems as they do not respond to the changes in pressure.

 

[cry22]

Morning SAK
Heard all the good about PICV, also heard about their prohibitive cost. Probably OK if only a few coils are involved, such as large AHU's. PICV's on a FCU system would cost an arm and a leg.
I do not see however any scientific data behind all the good that is claimed by PICV manufcaturers 9no statistical data available), especially in terms of savings, maintennace, etc.. actually, there is more maintenance involved.

Manual balancing valves are supposed to limit flow to no more than what the zone is supposed to get, the system response to pressure is addressed by the 2-way control valves, not by the Bal valves. In clear, I do not see the point against the manual balancing valves.

REM,
If only a few coils are involved, you could consider a circulating pump around the cooling coil between supply and return (similar to a freeze protection pump) to make sure that the return water out of the coil has the 10, 12 or 14 degrees delta T desired.

 

[SAK9]

Cry,

If a PICV is effective, you will get higher than design delta T across the AHU under low loads, a stable two way valve trend report and a stable temperature trend with in the space. All of these are easily verifiable.

My peeve against manual balancing valve is that they are set for a single static design condition whereas the system parameters are constantly changing in a variable flow system. So it is possible that AHUs near the plant may starve under certain light load conditions as they have been balanced to kill the excess pump pressure at design conditions (exclude cases where pump pressure is on auto reset)

 

[cry22]

SAK
In my opinion, Balancing valves are no more than shut-off valves partially open or closed, I do not see their imnpact on the system pressure at all.
I think that the savings or piece of mind that may be brought by PICV's are not worth the expense (we are talking about several hundred dollars on a typical building) for little or no gain to justify the cost.
I would rather spend that kind of money of something tangible, such as a grey water system, a solar panel system for domestic water, daylight harvesting system, Lutron light dimming systems, etc..

we are supposed to address as we say in the Commissioning world, "the Owner Project Requirments". I have yet to see an Owner calling for PICV's. The owner typically does not know and relies on the Hired expert.

it is my opinion that hired Experts should weight the cost benefits of any expense, especially when prohibitive such as PICV's.

I don't dislike PICV's. I have never heard anything bad about them, BUT, I have yet to see their money's worth neither. If anyone here has data, please share with us.

 

[Drazen]

cry,

static balancing valves in variable flow systems will cause erratic regime, and such systems need differential pressure regulators as a must, so in cost analysis you need to take them into account, as well as the fact that on larger system one regulator cannot give satisfactory results for the whole branch, so you need to make subbranches and take that into account as well.

i actually consider picv cheaper option.

on the other hand, redesigning non-functional static systems gives me some more job.

of course that does not mean that i would ever misguide somebody just to get more job, but it does mean that some of stubborn "old-school"-ers insist on ignoring some of recommendations provided to them, and than go for it afterwards.

 

[mauricestoker]

I believe the same arguments have been made before regarding pressure dependent vs. pressure independent mixing boxes.

All cooling coils where I work include requirement for PICV, which is also in the OPR for the CxA.

If we were going to keep the pneumatic controls, never use a VFD, and keep the inlet guide vanes on units, we probably would not select PICV's. If this were for just a few coils, a 3-way would be better. I can't imagine operating a campus distribution variable primary chiller plant or distributed secondary without PICV's.

 

[marcoh]

My guess is that the AHU's at the ends of the run are 3 way valves and not 2 way valves.

Some people include 3 way valves at the end of runs to ensure cold chilled water is always available, but this results in low return water temperatures if this AHU is not at 100% cooling. My preference is to only install 2 way valves, and if the end AHU starts and requires cooling, waiting 15 min for cold chilled water to arrive is unlikely to be an issue.