Repository with code to
- build a database with relevant data from public Heatpump Keymark Datasets.
- identify efficiency parameters from the database with a least-square regression model, comparable to Schwamberger [1].
- simulate heat pump efficiency (COP) as well as electrical (P_el) & thermal power (P_th) and massflow (m_dot) as time series.
For the simulation, it is possible to calculate outputs of a specific manufacturer + model or alternatively for one of 6 different generic heat pump types.
[1] K. Schwamberger: „Modellbildung und Regelung von Gebäudeheizungsanlagen mit Wärmepumpen“, VDI Verlag, Düsseldorf, Fortschrittsberichte VDI Reihe 6 Nr. 263, 1991.
For reference purposes:
- DOI:
- Citation: Tjarko Tjaden, Hauke Hoops, Kai Rösken. (2021). RE-Lab-Projects/hplib: heat pump library (v2.0). Zenodo. https://doi.org/10.5281/zenodo.5521597
If you're interested in how the database and parameters were calclulated, have a look into the Documentation HTML or Jupyter-Notebook. There you also find a simulation examples and a validation.
The hplib_database.csv contains the following number of heat pump models, sorted by Group ID
| [Group ID]: Count | Regulated | On-Off |
|---|---|---|
| Outdoor Air / Water | [1]: 5812 | [4]: 40 |
| Brine / Water | [2]: 283 | [5]: 194 |
| Water / Water | [3]: 6 | [6]: 6 |
All resulting database CSV file are under 
The following columns are available for every heat pump of this library
| Column | Description | Comment |
|---|---|---|
| Manufacturer | Name of the manufacturer | 30 manufacturers |
| Model | Name of the heat pump model | 506 models |
| Titel | Name of the heat pump submodel | use titel name for simulating |
| Date | heat pump certification date | 2016-07-27 to 2021-03-10 |
| Type | Type of heat pump model | Outdoor Air/Water, Brine/Water, Water/Water |
| Subtype | Subtype of heat pump model | On-Off, Regulated |
| Group ID | ID for combination of type and subtype | 1 - 6 |
| Rated Power low T [kW] | Rated Power for low temperature level | -7/34 °C |
| Rated Power medium T [kW] | Rated Power for medium temperature level | -7/52 °C |
| Refrigerant | Refrigerant Type | R134a, R290, R32, R407c, R410a, other |
| Mass of Refrigerant [kg] | Mass of Refrigerant | 0.15 to 17.5 kg |
| SPL indoor [dBA] | Sound emissions indoor | 15 - 68 dBA |
| SPL outdoor [dBA] | Sound emissions outdoor | 33 - 78 dBA |
| Bivalence temperature [°C] | Minimum temperature heat pump is running without supplementary heater | *T_biv not used in simulation |
| Tolerance temperature [°C] | Minimum temperature heat pump is running with supplementary heater | *TOL not used in simulation |
| Max. water heating temperature [°C] | Maximum heating temperature | *T_max not used in simulation |
| Poff [W] | Eletrical power consumption, ? | *P_off not used in simulation (0-110 W) |
| PTOS [W] | Eletrical power consumption, ? | *P_tos not used in simulation (0-404 W) |
| PSB [W] | Eletrical power consumption, standby mode | *P_sb not used in simulation (0-110 W) |
| PCKS [W] | Eletrical power consumption, ? | *P_cks not used in simulation (0-99 W) |
| eta low T [%] | Efficiency for low temperature level | 105-300% |
| eta medium T [%] | Efficiency for medium temperature level | 107-202% |
| SCOP | seasonal COP | 2,7-7,7 |
| SEER low T | seasonal EER for low Temperature Level | 3,39-12,93 |
| SEER medium T | seasonal EER for medium Temperature Level | 5,04-13,87 |
| P_th_h_ref [W] | Thermal heating power at -7°C / 52°C | 2400 to 69880 W |
| P_th_c_ref [W] | Thermal cooling power at ? | 3000 to 53200 W |
| P_el_h_ref [W] | Electrical power at -7°C / 52°C | 881 to 29355 W |
| P_el_c_ref [W] | Electrical power at ? | 881 to 17647 W |
| COP_ref | COP at -7°C / 52°C | 1,53 to 7,95 |
| EER_ref | Electrical power at ? | 1,99 to 10,8 |
| p1-p4_P_th | Fit-Parameters for thermal power | - |
| p1-p4_P_el | Fit-Parameters for electricl power | P_el = P_el_ref * (p1T_in + p2T_out + p3 + p4*T_amb) |
| p1-p4_COP | Fit-Parameters for COP | COP = p1T_in + p2T_out + p3 + p4*T_amb |
| MAPE_P_th | mean absolute percentage error for coefficient of performance (simulation vs. measurement) | average = 19,7 % |
| MAPE_P_el | mean absolute percentage error for electrical input power (simulation vs. measurement) | average = 16,3 % |
| MAPE_COP | mean absolute percentage error for thermal input power (simulation vs. measurement) | average = 9,8 % |
| MAPE_P_dc | mean absolute percentage error for coefficient of performance (simulation vs. measurement) | average = 19,7 % |
| MAPE_P_el | mean absolute percentage error for electrical input power (simulation vs. measurement) | average = 16,3 % |
| MAPE_EER | mean absolute percentage error for electrical input power (simulation vs. measurement) | average = 16,3 % |
- Get repository with pip:
pip install hplib
or:
- Download or clone repository:
git clone https://github.com/RE-Lab-Projects/hplib.git- Create the environment:
conda env create --name hplib --file requirements.txt
Create some code with from hplib import hplib and use the included functions hplib.load_database(), hplib.get_parameters, hplib.HeatPump(), hplib.HeatPump.simulate(), hplib.HeatingSystem.calc_brine_temp() and hplib.HeatingSystem.calc_heating_dist_temp()
Hint: The csv files in the output folder are for documentation and validation purpose. The code and database files, which are meant to be used for simulations, are located in the hplib folder.
The European Heat Pump Association (EHPA) hosts a website with the results of laboratory measurements from the keymark certification process. For every heat pump model a pdf file can be downloaded from https://keymark.eu/en/products/heatpumps/certified-products.
This repository is based on all pdf files that were download for every manufacturer on 2023-04-17.
If you find errors or are interested in developing together on the heat pump library, please create an ISSUE and/or FORK this repository and create a PULL REQUEST.
MIT License
Copyright (c) 2023
You should have received a copy of the MIT License along with this program. If not, see https://opensource.org/licenses/MIT
