We do all our controls on the BMS, and trust me, just about any BMS system is fast enough for chiller plant optimization. It is not a high-speed system, where you need thousands of computations a second. we run our reset schemes on 5 second intervals, and even that is excessive. The important part is that it is designed to be flexible, and let the system be optimized over time. I understand there is a need to get the system up and running simply, to complete the construction phase, but there should be included a post-opening/post-startup period of trending, analysis, reconfiguration, trending, rinse-repeat.
Originally Posted by motoguy128
As for running a chiller in free-cooling, that is usually not worth the effort, and it is much simpler/cheaper to run a plate-frame exchanger. A better opportunity is to run the plate and frame as its own primary loop from the chilled water return, and then through the chillers. Great for those situations where you need a 44°CHWS, and have a 45°F tower water. It allows you to get as much of the free cooling direct from the condenser loop, and then top the cooling off with the chiller. Best of all, it's not hard to add to an existing plant, as long as you can find space for the heat exchanger and its own primary pump.
Great idea on the he. however I. One area with most of our chillers, we run glycol with a setpoint of 0f. its p.s piping to a big storage tank without much stratification. So at low loads in winter I think the evap return is only 3f.
Id rather use a bms like tracer summit with preconfigured interfacers for trane chillers, lead lag schemes, etc. We have foxboro dcs. It configured to control most things with 10 to 100 second intervals. Mostly slow processes. It not the speed, but that pur controls enginers are nog used to this type of equipment. The current chiller stages on rla and tends to load multiple chillers evenly and does prioritize running the centrifugals over the rotaries. The are times where a single cvhe could hold the load but instread 2 rotaries are also partly loaded.
You cannot remotely view approaches, pressures or any useful info. They even wasted money on seperate high dollar rosemount transmitters for. Water temps.
Sent from my SGPT12 using Tapatalk 2
One consideration to CHW temperature reset is that the lower the setpoint the more latent heat will be absorbed by the CHW loop and this will add tonnage to the chiller.
This is only indirectly true. If the air handler is controlling to a discharge air temp, and is modulating the chilled water valve, then changing the chilled water temp will not affect the discharge air temp - which equates directly to the amount of moisture removed. If you have an air handler set to discharge 55°F air, and dewpoint of the incoming air is higher than that, then the amount of latent heat removed is going to be the same no matter what temp the chilled water is going in.
Originally Posted by HVACControlTech
That's not to say that latent heat removal isn't to be considered, but that is an air handler optimization, not a chiller plant optimization. Ideally, you'd run the discharge temperature only as low as the building needs to meet desired indoor temperature and humidity.
I've seen kW/Ton Plant wide vary from 1.23kW/Ton to 0.83kW/ton. Sometimes district loops play a factor.
Improvements vary from 50% to 20% reduction - and that pertains to some serious plant optimization
strategies (eg, Hartman Loop).
Then there's just the 'politics' of the matter: what do you think saving ~300K/year means to a $4B biotech company?They just bought the optimization product to say they wave the 'green flag'.