Dear All,
I am comparing the energy performance of a constant volume and variable volume
system for a composite climate of India (New Delhi) in EnergyPlus. The building
area is 7500 sq m. The chiller capacity is same for both the cases. The chilled
water to the cooling coils is supplied by a constant speed pump.
The savings in the fan energy is evident due to variable speed of the supply fan
in case of VAV. However i am getting energy saving in cooling energy (chiller
energy) also, in VAV compared to CAV, which i feel is due to higher delta T
(chilled water) across cooling coil for CAV compared to VAV, this in turns
increases the chiller electricity consumption. However while discussions with
some consultant, it is their feeling that the cooling energy would remain same
for both type of systems.
I want to ask has someone else tried this simulation and would there be any
difference in cooling energy or not?
Thanks in advance.
Deepak
Deepak,
Here are some thoughts:
? Less fan energy = less cooling load, since the fan energy is a part of the total cooling load.
? If the pump is variable volume, the pump energy required for VAV fans will be slightly less due to less fan heat to cool.
? If you are using the identical chiller for each system (VAV, CAV), then the chiller should use less energy also, due to less fan and pump heat.
? A more common comparison would be to contrast a VAV chiller system with a CAV packaged rooftop system. For that comparison, the part load efficiencies of chiller and compressor / DX coil will be a major factor. Dehumidification will also be different for DX vs. chilled water coils.
James V. Dirkes II, P.E., LEED AP
There are several things going on here:
- Yes, the lower fan energy for VAV will add less heat to the chilled water
loop. This will propogate to the pumps, chillers, and heat rejection.
- Chilled water coils are complex heat-exchangers, which do not exibit
linear behavior. Reducing the air flow decreases the water-side coil
effectiveness; therefore, at a given load, a CAV system will have a larger
chilled water delta-T.
- With equivalent pumping schemes, this will result in pump savings for the
CAV system (in the absence of 3-way valves). I'm not sure about ePlus, but
this can be demonstrated in the latest DOE2.2.
- In reality, the larger chilled water delta-T will decrease chiller lift
and increase its efficiency. Again I'm not sure about ePlus, but in DOE2.2,
chiller curves are a function of a dT parameter which is the difference
between condenser entering and chilled water leaving temperatures. Its an
attempt to account for chiller lift, but it does not give an efficiency
credit for increasing the chilled water delta-T. So, you will not see the
chiller efficiency boost in DOE2 for CAV systems due to a greater chilled
water delta-T.
In my experience, the VAV fan savings (and reduced chilled water load
savings) usually outweigh the pump and chiller savings for CAV. However,
it varies from building to building. For example, if you had a rare
building with a low air-side static pressure drop to begin with and a long,
high head pumping system, then its possible that the CAV system will be more
efficient overall.
Aaron
I have a bit different view of this:
The load seen by the chiller is not just the load in the zones, but the load
from the system. The system load is what is required to cool the supply
flow down to the supply air temperature (SAT) setpoint. The SAT is either
fixed, or cold enough to satisfy the zone requiring the most cooling (the
"critical" zone), assuming the zone loads are met. All zones except for the
critical zone don't need as much cooling. In a VAV system, the cooling is
reduced via reduced flow. This reduced flow means less system load to the
chiller plant. In contrast, the CAV system load does not get reduced
because some zones have lower loads. The full fan flow needs to be cooled
to the SAT. In this case, the non-critical zones are kept from being
overcooled by the addition of reheat.
Regards,
Bill
Aaron
Can you elaborate how the larger chilled water delta-T will decrease chiller
lift and increase efficiency?
Bill, you're right about that. A central CAV system with reheat is limited
by the most loaded zone. My statements before were simply for one air path
and one zone, which is probably rare. The degree to which this hurts a CAV
system depends on the diversity of the loads. If all zone loads remain
similar, then the supply air temperature off the CAV coils will start to
float upwards at reduced loads, and you'll see an increased chilled water
delta-T.
As for the increased chiller efficiency, here's my thinking. At a constant
load with no mixing valves, an increased delta T across the cooling coils
allows for a lower gpm. Slowing the water flow rate and raising the
entering water temperature increases the effectiveness of the evaporator
barrel (i.e. the enthalpy of the refrigerant now has a greater ability to
approach the enthalpy of the higher enthalpy water). Running the compressor
as before would over-cool the water. To maintain a constant chilled water
supply temperature, the mass flow of the refrigerant must be reduced.
Depending on the type of compressor unloading mechanism, this should produce
some energy savings.
This is my understanding simply based on theory, and I'm sure it's much more
complicated in reality. Again, I know that DOE2 does not account for
varying inlet conditions to the chiller evaporator barrel, but I'm not sure
about ePlus.
Aaron
I too cannot claim to be capable of tearing a chiller apart and putting it back together, but I can affirm Aaron's fundamentals appear on-cue to my understanding. Less cooling ? less compressor work ? $$$ savings.
The attached email exchange from a few months back may be of interest to those following/participating in this discussion, though it is a bit lengthy. In it, I attempt to "reverse engineer" the three default curves found in eQuest/DOE2 which together define behavior of a centrifugal chiller, observing and discussing the isolated effects of all the variables that are taken into account on an hourly basis... One following along can see I was a bit puzzled along the way, but arrived at an "ah-ha" moment when I followed through and applied the curves against each other to observe their net effects on efficiency under set conditions.
My hope then and now is that sharing this thought process may help others arrive at their own "eureka" moments =).
NICK CATON, E.I.T.