EnergyPlus HVAC system models are scalable. What this means is that for a specific hardware configuration and associated performance, the model assumes a different size system would perform similarly (i.e., use the same performance curves or input criteria). This is not always true in real life where a range of HVAC system sizes uses the same chassis leading to slight differences in performance as capacity changes through the manufacturer line of systems using this chassis.

To use the scalable sizing methodology, use autosize for all inputs. This method will scale the HVAC system to meet the loads in the simulated building. If you want to model a specific HVAC system for a specific building, then enter the actual capacity and air flow (and other information as needed). If you know only the capacity, search for manufacturers data for that specific HVAC model to fill in the blanks (air flow, SHR, COP). If you can not find manufacturers data use rule of thumb data. For example, if you have capacity and not air flow, use air flow = capacity * 0.00054 (or about 400 cfm/ton which is typical of DX equipment), SHR = 0.78 and COP = ~3.5. Of course, these numbers do vary with different HVAC systems.

If this air flow ( capacity * 0.00054 ) is much different than that predicted by EnergyPlus, then either the building loads are not similar to the real building (i.e., the building or internal loads are different) or the Sizing* objects do not reflect the correct information.

The bottom line is that your choices are to fully autosize the HVAC system or enter all values for autosizable fields. It is very risky to autosize some fields and enter values for other autosizable fields.

Running your file in V8.7 shows several warnings like this:

```
** Warning ** ProcessScheduleInput: Schedule:Compact="OCUP_OFICINA" has missing day types in Through=12/31
** ~~~ ** Last "For" field=FOR: WEEKENDS HOLIDAY
** ~~~ ** Missing day types="SummerDesignDay","WinterDesignDay","CustomDay1","CustomDay2"
** ~~~ ** Missing day types will have 0.0 as Schedule Values
```

This can lead to incorrect HVAC system sizing, more so for the summer period since internal gains add to the load.

The eio file shows air flow to capacity ratio = 0.0000604 which is at the high end of the range of allowed air flow ratios. This means that the calculated zone air flow is high compared to the loads and the DX coil model increased the system capacity to stay within a predetermined air flow per capacity range.

The zone sizing results (*zsz.csv) shows why the top end of the air flow to capacity range was used, where cooling dominates and peak air flow to capacity = 2.98/24900 = 0.00012 (about 900 cfm/ton which is way too high). I should mention that I used Chicago weather so your numbers will be different. So if you enter these values directly into the DX coil, the errors tell you that the capacity and air flow are not within a range typical ... (more)

What exactly are you changing in the model you linked to get this error? I assume your hard sizing the capacity and SHR.

At first, I run model with

autosizefor:I get 12,2 kW, 2.57 m^3/h for heating, and 25,9 kW, 0.34 m^3/h for cooilng, in EIO file. If I change autosize fields for this values, I get error.

I'm not sure why you'd want to do that, but try hardsizing SHR too and see if the error goes away.

I get this error

.Min Rated Vol Flow Per Watt=[4.027E-005], Rated Vol Flow Per Watt=[1.031E-004], Max Rated Vol Flow Per Watt=[6.041E-005]. See Input Output Reference Manual for valid range.What I'd like to do is to simulate a real system, in a real building. I know how to use autosize; reading EIO file, I know what capacity and equiment I need. However, if I'm analisying, a real heat pump in an existing building, I dont know how to simulate this system in my model.

Could you modify the model to incorporate an 10 kW heat pump, for example, to know how the bulding is responding?

That's a pretty high air flow rate for DX equipment which usually operate at around 400 cfm/ton. Your value of 0.0001031 equates to about 770 cfm/ton, where did that number come from? What are your inputs for capacity and air flow? If you divide air flow by capacity, to get flow per watt, you should be somewhere between 0.00004 and 0.00006.