Chiller Capacity - Measureing loss of capacity

[TECMSC]

Good Day,

I'm trying to figure out an accurate method of measuring the loss of air cooled chiller capacity when outside air temp. is above design criteria.

the goal: at today's heat load and chiller loading, what outside air temp would we need to bring on another chiller?

Initially i was on the path of using volumetric eficiency. I have an "old" Trane table that approximates efficiencies of R-22 reciprocating compressors at various compression ratios. Using volumetric effeciency(from Trane table) and compression ratio (discharge psia/suction psia) i could calculate the pounds of refrigerant circulated. info required to use this method would be refrigeranat, bore and stroke of compressor, # of cylinders, evaporator and condenser temperature at design condition. From this I could determine the theoretical and actual weight of refrigerant pumped by the compressor and corresponding theoritical and actual capacity (tons). I could plot the various compressor capacities (tons) at increasing condensing temps. Once the calculated tonnage is greater then installed, we would need another chiller. Question: would this approach work with refrigerant other than R-22 and recip. compressors i.e. rotary? Could I assume the suction pressure remains constant while the condensing pressure increases?

After considering this approach I ask -- Is there another method to calculate capacity loss?

Another, less time consuming approach could be to get a compressor rating table specifically for the chiller compressors. I could compare the capacity in tons at various operating conditions i.e. suction and condensing temperature and get the manufacturers rating (tons).
Question: would I only need to get the compressor data sheet or just use the chiller unit capacity ratings (tons) at different outside air temps.?

thanks for any comments/suggestions in advance.

 

[KiwiMace]

Why not just monitor the Compressor power? This is the limiting factor regardless of what the chiller is actually doing w.r.t. fouling, high ambient, lowered chilled water supply temp. etc etc.

To measure tonnage - calc with the flow and entering/leaving conditions and you will also get a bonus efficiency calc.

 

[TECMSC]

You're suggesting:
-Measure compressor amp draw and convert to horsepower.
-Record tonnage based on CHW flow.
-Result is HP/ton at current ambient and internal loads.

How do I predict the refrigeration tonnage capacity at 115degF ambient with this info?

 

[sterl]

Recips are automatic valve machines: In good condition, their net displacement is predictable just as you indicated....

Rotary takes in a little more...

Screws and rotary vanes are largely Fixed - Volume Ratio machines....Pressure Ratio and polytropic efficacy can be used to predict mass flow, as the rating programs of a variety of manufacturers do, for any set of conditions short of the ridiculous.

Scrolls are a little more of a wild card because of the "compliance" topic but they will play by similar rules....

Turbos...Well, there's a bunch of design parameters there, you would either have to know the iso-poly model (as per the maufacturers ratings) or you would need to re-characterize the Flow-Head curves as polytropics.

That last sounds like a lot of fun, but not for me, today.

 

[absrbrtek]

I would contact the manufacturer to get the performance data on your specific chiller at various ambients and loads. I am sure they could plot 115F ambient for you and determine the capacity under those conditions. They should all have design software to due this with minamal guess work involved.

Other than that, write a program using cw outlet temp degree minutes starting the next chiller when needed. Take in to consideration time delays etc. This would also automaticaly start the next chiller should the previous chiller have a failure, rather than just depending on outside ambient for this function.

 

[KiwiMace]

TECMSC - you mentioned that your goal was to understand what outdoor air temp to bring on another chiller. What I was suggesting was that when your chiller amps are max, then you need to bring on another chiller.

Chiller amps are the product of a mystical formula that is proportional to actual delivered tonnage with all external driving conditions, efficiencies, inefficiencies, malfunctioning systems, etc.

Look at steady state amps drawn for a given tonnage delivered on similar weather days for different years. As your amps go up for say, 100RT delivered then you can determine that your capacity has degraded.

 

[AbbyNormal]

Abs, I still remember an eight your old lesson on Farhenheight Seconds from you. How about expanding the degree minute concept?

Take the "V" out of HVAC and you are left with a HAC(k) job.

 

[AbbyNormal]

lol "8 year old"

Take the "V" out of HVAC and you are left with a HAC(k) job.

 

[TECMSC]

...been traveling with limitied internet acess...

CinceMace-thanks for the input...I now understand what you were getting at

Absrbretk - this is an air cooled chiller, not sure if your approach would work-maybe measure condenser airflow temps?? or condenser pressures?? -- besides, you would have to expand on the degree minutes concept...[i don't think i could write a program]...thanks

Sterl-thanks as well, this gives me another perspective to help understand the big picture.

Abby - are you serious on the Farhenheight seconds?

I think I'll have to keep it simple (KISS)i.e. using manufacturing charts and actual data(amperage and flows)

 

[AbbyNormal]

yes the F-seconds was a low limit set up. Similar to "degree days" it was the product of a time in seconds and the degrees below a low limit set point.

Something very low below a limit would trip out after a short time, something that was marginally low would trip out after a long time period.

So I an not familiar with how the 'degree minutes' worked so I asked the question.

Take the "V" out of HVAC and you are left with a HAC(k) job.