OpenStudio Chiller Comparison - Troubleshoot Curves

In this video we will discuss how to troubleshoot the biquadratic and quadratic characterization curves. We will discuss some common mistakes and briefly describe how the curves are used by EnergyPlus. Finally, we will run the simulation to quantify energy savings by replacing the older chillers with new.

Transcript:
Okay. It was successful. Let us go to our error file and see if there was any warnings generated.
Okay. This is what I suspected. It is saying that our capacity ratio as a function of temperature 
curve is not equal to one at the rated conditions.
It is also saying the energy input as a function of part load ratio (PLR) curve is also not equal to 1 at the rated conditions.
It is repeating the warning for the other two chillers. We have two curves that we need to take a look at a little bit closer.
Let us first go to the capacity as a function of temperature curve. Let us see.
Capacity as a function of temperature. These values are in SI units, so the temperatures are in celsius.
We can go back to our model. Go to the HVAC tab.
We will take a look at our chilled water loop. The chillers. These are all in IP units.
We can quickly switch back to metric units by going to Preferences, Units, Metric.
Now we can see what our design our reference conditions are. Our reference conditions for this chiller are approximately 4.5 celsius evaporator and 26.6 celsius condenser. ~4.5°C and ~27°C. 
You can see the output of our curve is 0.65 at our design conditions. You can see that in the the output file.
It is 0.653 for the output of that curve. At our design conditions this value should be 1.
This value, 1, gets multiplied by our reference capacity. Our reference capacity was 14,208 kBtu/hr (4.16kW) at design conditions. So, 14,208 kBtu/hr x 1 (design conditions) is 14,208 kBtu/hr.
So, this is a problem. Similarly, if we took a look at efficiency, the efficiency at design conditions should be 1 as well. It is pretty close 0.99.
You can see that by the output of our curves. This mathematical curve fits the data by 92%. That means it is a pretty good curve.
Unfortunately, this curve is only at about 16% fit. 16% fit for capacity.
Back to capacity, you can see that it is way off. It should be closer to 1. So, there might be some 
problems with this.
We can immediately recognize that the curve dips down to zero capacity for a low condenser temperature and a high chilled water temperature. That does not make sense.
If your condenser temperature is low and your chilled water temperature is high, you should have the most capacity from the chiller.
This curve really should slope down from somewhere close to 1 up in this corner all the way down to this corner here.
We are missing some data that creates this curve.
We can take a look at the numbers that we input. We have a lot of good data for 40°F (4.4°C) chilled water temperature.
We have mostly variable data for the condenser temperature.
You can see that the chilled water temperature is not variable. It is all 40°F (4.4°C). So, we are missing some information here.
If we look at the chiller performance...it is what we had stated for our boundary conditions...our design chilled water temperature is 40°F (4.4°C).
It is plus or minus 5°F (2.7°C). This range of chilled waters should go from 35°F (1.7°C) to 40°F (4.4°C) to 45°F (7.2°C). The data that we input only only goes to 40°F (4.4°C). It is very sparse in the high temperatures.
We do not have any temperatures that go down to 35°F (1.7°C).
That is some data that we need to ask the manufacturer about.
The other thing; you will notice our design conditions are 40°F (4.4°C) and 80°F (26.7°C).
We have a fixed capacity and a fixed energy input for that value.
But,we have a lot of other values at 40 and 80 here. These represent different part load ratios (PLRs) as the chiller goes from 100% down to minimum part load.
All of the values on this table should be at a 100% PLR for the reference conditions.
We need to get rid of some of these lower PLR values. The mathematical curve output is at 100% PLR.
That 100% PLR gets multiplied by the capacity to give you multiple steps all the way down to the minimum capacity.
This curve really should be somewhat flat depending on the chiller characteristics. It should not have that much slope to it.
As you reduce your PLR, it steps down this this flat surface, it gets lowered down further and further based on its PLR.
This curve should be generated at 100% PLR.
If we go back to our chiller performance...we have a whole bunch of extra PLRs stuck in here.
That is another thing we have to go back to our salesman/equipment manufacturer and get more information to fill in some of these values.
We need more data for low chilled water temperatures, in this regime, for a range of condenser temperatures. We are missing data on this side of the curve.
...
Full transcript at:https://www.helix-engineers.net/techn...