Suggestions on how to size mutilple chiller plant 2

[UtilityLouie]

I have a new application. It is essentially a copy of another installation in the same geographic area.

In the original installation some assumptions were made with regards to heat load from electrical gear. The actual chiller loads we have been seeing are about 70% of what we thought design was based on trend data.

My issue is: In the new installation, I want to size the new chillers closer to the load. I am going to have one absorption unit and one centrifugal unit with a VFD.

The load break down is (in percentage of operating time)

200-300 tons - 54%
300-324 tons - 46%
325-349 tons - 25%
350-374 tons - 13%
375-399 tons - 6%
400-424 tons - 3%
425-449 tons - 1%
450-474 tons - 1%
475-499 tons - <1%
Over 500 tons - <1% (Peak tonnage was 765)

The original system was two 500 ton chillers. My thoughts are - sizing the absorption unit at 350 tons and operating it as the primary chiller. This size is due to problems with absorption unit turn down, I think it should be running as close to full load as possible.

My gut tells me that with a VFD, I can size the centrifugal as big as I want to cover my butt for back up operation. But my experience with chilled water systems says - "Don't be too hasty Grasshopper".

Do I size them both at the same tonnage? Suggestions?

In this new engineering world we live in (i.e. me being relatively inexperienced with no knowledgeable mentor where I work - just non-detailed manager types) this forum is awesome for checking out ideas.

 

[imok2]

When energy conservation is a consideration it’s recommended that you
size multiple chiller systems differently - Most installations having a chiller plant should have multiple, unequal size chillers. This allows for the most efficient chiller for a given load to operate and I agree with your decision to run the absorber as fully loaded as possible as it should increase it’s performance. Any time you can balance the load on the chiller to run at 90% + it results in better KW/ton.

 

[quark]

Can you have a look into your figures? The total time is morethan 100%.

Anyhow, I agree with both of you on running the absorption system continuously. But you should take care that absorption units are much prone to derating over a period of time. My machines were giving me 80% of designed capacity after 4 years of operation, irrespective of correct parameters of operation and religious yearly cleaning of LiBr and the system.

If I go with a 350TR VAM, centrifugal chiller may have to run at 5% capacity with one increment in load pattern. Since 350-374 TR load occurs for 13% of time, you will be running your centrifugal chiller at lowest efficiency for the highest duration of time (Mind that 5% is just the lower most boundary of a centrifugal chiller capacity control). I have seen manufacturers showing good efficiencies of centrifugals even at 5% load by dual capacity control (guide vane + VSD), I am doubtful as to how far we can maintain the same conditions as they do during testing.

In any case, you have a load increment of 25TR and the centrifugal has to run at 5% load, I would reduce the period of operation.

As you mentioned in a previous thread that you have free steam, I would very well go for a 400TR machine so that I may have to run the centrifugal chiller at the lowest efficiency for a mere 3% of operating time.

I strongly believe (and roughly calculate) that if you go with thermal storage systems with a 400TR VAM, you can have your centrifugal chiller as 100% stand by.

 

[UtilityLouie]

Quark,

Thanks for your input. I have revised my numbers. I didn't do the sanity check until after I posted. Here are the revised numbers. (I've added some other 25T bins after I realized my mistake.)

200T-224T: 3%
225T-249T: 5%
250T-274T: 12%
275T-299T: 27%
300T-324T: 21%
325T-349T: 12%
350T-374T: 6%
375T-399T: 3%
400T-424T: 2%
Over 425T: 1%

The remaining 7-8% is either data that I threw out because our DDC system wasn't working properly, or the system was shut down.

I only dream about enough square footage (and budget) for a thermal storage system!

This new (Correct) data seems to skew things more toward a 375T unit as the the primary. Do you think picking the centrifugal to be 375T with a VSD, vanes and hot gas bypass would be the best option? It is definitely the least creative option... The chillers would then be able to operate at part load during peak conditions - not the most efficient, but with this equipment in my industry - the operators want easy not efficient. The managers want efficient, and if it's hard to operate, too bad. Operations and maintenance staff are the first to go.

 

[quark]

I would choose both the machines with equal capacity. By this, you will have 100% backup atleast for 80% of time.

I am unaware of hot gas bypass for a centrifugal chiller. It is OK if this is used as a surge control mechanism, otherwise this is a wastage of energy as far as a centrifugal chiller is concerned.

As far as my experience is concerned, VSD supplemented by guide vane control does good capacity control even at very low loads. Of all the options available, you don't have any other machine that can give you better efficiency at low load condition than a centrifugal chiller with VSD. So, this may be your obvious choice.

 

[Yorkman]

Quark,
I'm not pushing York product I'm just more familiar with it. A hotgas bypass option can be ordered with the YK centrifugal machine (R-134a) they also have a variable geometry diffusor, fancy talk for a moveable diffuser wall. Both these options allow running the machine under load condtions that you never could before.

I'm not a real engineer, but I play one on T.V.
A.J. Gest, York Int./JCI

 

[imok2]

Yorkman I believe Mcquay has the same thing, a great concept.

 

[HVAC68]

With so much of part load numbers and that at varying percentages, my suggestion would be to go in for screw chillers.

HVAC68

 

[quark]

Yorkman,

I had experience with R123 500TR (x3) millenium chillers of York and absolutely good w.r.to performance and reliability. Moreover, York guys are good and give fast response. Trane is aggressively marketing for screws in India and somehow are reluctant to push centrifugals.

HVAC68,

My observation and study indicates screws are no better for loads beyond 300TR. The only advantage I see is about the multiple compressors. Apart from the metallic noise, it is not totally a stepless control. Can you indicate me some advantages of going for screws for above mentioned loads?

Regards,

 

[HVAC68]

Multiple compressors are one of the advantages.
Slide valve control is available with most of the manufacturers. Carrier, which didn't have the same, has also introduced now.
VFD screws are also available today.
From a maintenance perspective, with hardly any "touching" parts, screw is a better one compared to centrifugal for loads upto 300 to 350 TR, especially with more part-load requirement.
As regards the noise - this can be attenuated and is probably not the biggest of problems.

HVAC68