Frosting Penalty | Knowledge Base

When the HP operates during low ambient temperatures (< ~9°C), frost will begin to form on the HP evaporator coils. This frost blocks airflow, reducing the efficiency (COP) of the HP.

The COP and Power coefficients derived from AS/NZS 5125 regression only include data from test conditions 1-4. This is because not all heat pumps are designed to operate in low temperatures, some will fail at cold temperatures, the element will be used instead.

To calibrate the model for low temperature performance, we need to compare the COP tested at 1°C against the COP simulated in TRNSYS. This percentage difference is called the Frosting Penalty, and will be applied when the dry bulb temperature is lower than initial frosting temperature (Tifr).

Recall Tifr = initial frosting temperature = the lowest ambient air temperature (dry bulb temperature) in test condition 1. It is usually in the range of 7-9°C.

This is known because in test condition 1, the test engineer will try to run the test at the lowest dry bulb temperature possible that does not produce any frosting. If they see frosting, they have to retest at a slightly higher dry bulb temperature until no frosting is produced.

The frosting penalty, or ‘low temperature operation penalty’, is a modelling parameter that calibrates the model against the AS/NZS 5125.1 low temperature test.

To compare the COP % difference, we create a TRNSYS model that mimics the 5125 test conditions: same water temperature, same dry/wet bulb as in the test.

Then we simulate a single heat-up cycle, taking the tank from fully cold to fully hot - roughly 10°C to 60°C water temperature.

The TRNSYS result will be over predicting the COP on average, because it is not taking into account the effects of frosting, which reduce the COP by 10-50%.

So we need to calculate the percentage difference between the simulated COP and the tested COP. This percentage difference is known as the Frosting Penalty.

The frosting penalty is calculated from the following equation:

FP = (1-COP_5125/COP_TRNSYS)

Once the frosting penalty is applied, the TRNSYS model will now predict the same average COP as in the test.

The Frosting Penalty reduces the COP to match the tested COP at 1°C, and TRNSYS linearly interpolates between 1°C and Tifr .

COP' = (1-FP)*COP

The frosting penalty is only applied in TRNSYS when the ambient temperature is below the initial frosting temperature (Tifr).

Low temperature penalty from AS/NZS 4234

Please now review AS/NZS 4234:2021 4.8.6.2.1 Low temperature operation penalty calculation.

Worked Example

A 250L integral HP model is tested to AS/NZS 5125 at 1°C. During the test, the HP frosted up 5 times, requiring several short defrost cycles, which reduce the efficiency. The average COP over the tested heatup cycle is 2.8.

A TRNSYS model of the 250L HP is created to mimic the low temperature test - same starting water temperature and dry/wet bulb temperatures as in AS/NZS 5125 real test. The HP efficiency is defined by the COP & Power coefficients which come from analysing test conditions 1-4, which excludes the low temperature test data.

When the TRNSYS model simulates the heat-up cycle at 1°C, the performance is over-estimated, as it currently does not take into account any frosting. The average COP over the simulated heatup cycle is 3.2.

How much higher is this compared with the test?

It is 3.2/2.8 = 1.14 => TRNSYS predicts average COP of 14% higher than 5125.

So how to calculate the Frosting Penalty now? We need to bring the TRNSYS average COP back down to the 5125 result.

FP = (1-COP_5125/COP_TRNSYS)
FP = (1-2.8/3.2) = 0.125 = 12.5%
The frosting penalty for this HP is 12.5%.

TRNSYS Tool Methodology

The TRNSYS Tool simulates a heat up at the same initial water and ambient air temperature as the 5125 test.

Follow these steps to calculate FP:

Data Preparation:
- Copy the 1-minute data from the low-temperature test into the “FP_DATA” sheet.
- Arrange columns in the following order: “Date,” “Time,” dry bulb (“DB”), wet bulb (“WB”), electrical energy (“E”), temperature sensors (“TC1,” “TC2,” …, “TC6”), and average temperature (“TAVE”).

If you ever need to calculate wet bulb from dry bulb & dew point, use this psychometrics calculator.

Rename columns to text in commas (i.e. Date, Time, E ect) to ensure proper identification by the macro.
Ensure that electrical energy is measured in watt-hours (WH).
If the column “TAVE” does not exist, create a new column: “TAVE” = average(TC1:TC6).
Trim the data to start just before a sensor reaches 10°C . In the screenshot below, that means deleting rows 2-10. This eliminates any startup effects.
- before one of sensors meets 10C

FAM_HP Sheet:
- In the “FAM_HP” sheet, locate and click the “Process FP data” macro button at the top of the sheet.
- Check the FP calibration rows for “Frosting penalty Calibration.”
Processed data input to calibration parameters in FAM_HP sheet
- Verify the details in the “FP_DATA_<ModelName>” sheet, ensuring that the calculation details are copied to the end of the list of sheets.
- Review all formulas to confirm they reference the correct columns and that the data is accurate.
Frosting Penalty Calculation:
- Click the “Calculate FP” macro button in the “FAM_HP” sheet to initiate the frosting penalty calculation.
- The resulting FP value should fall within the range of approximately 10-50%.
Frosting penalty = 20.1%
Verification:
- Check the “FP_<ModelName>” sheet for detailed calculations and confirm that the formulas reference the correct columns, ensuring data accuracy.
```
<figure><img src="/kms/image (54).png" alt=""><figcaption><p>FP_ECON155 sheet shows steps in frosting penalty calculation</p></figcaption></figure>
```

Video

Watch this video of a HP in defrost mode, completely covered in frost at first, which melts off over the course of about 5 minutes.