pyslpheat

Standard Load Profiles for Heat demand — Python implementation of two German heat demand profile standards for district heating simulations.

Module	Standard	Resolution	Method
bdew	BDEW/VKU/GEODE SigLinDe	Hourly	Temperature-continuous sigmoid + linear
vdi4655	VDI 4655	15-minute	Day-type step profiles

Installation

pip install .
# or editable install during development:
pip install -e .

Quick start

from pyslpheat import bdew_calculate, TRY_BAUTZEN_2015

# BDEW – hourly profile, 20 000 kWh/a, multi-family residential
# Uses the bundled Bautzen TRY (51.1676°N 14.4222°E, climate zone 9)
df = bdew_calculate(
    annual_heat_kWh = 20_000,
    profile_type    = "HMF",
    subtype         = "03",
    TRY_file_path   = TRY_BAUTZEN_2015,
    year            = 2026,
)
print(df["Q_total_kWh"].sum())   # → ~20 000 kWh
print(df["Q_total_kWh"].max())   # → peak load [kWh/h]

Both functions return a pandas.DataFrame with a DatetimeIndex and columns Q_heat_kWh, Q_dhw_kWh, Q_total_kWh, temperature_C (VDI 4655 additionally includes Q_electricity_kWh).

GUI – no scripting required

pyslpheat ships a desktop application that lets you configure all parameters interactively, inspect the resulting load profile in an embedded plot, and export results as CSV — without writing a single line of code.

Install with GUI dependencies:

pip install ".[gui]"

Launch:

pyslpheat-gui

The application has two tabs — one for each calculation standard:

Tab	Standard	Output resolution
BDEW SigLinDe	BDEW/VKU/GEODE SigLinDe	Hourly (8 760 values)
VDI 4655	VDI 4655	15-minute (35 040 values)

Both tabs offer the full parameter set including stochastic post-processing, discrete DHW draw events (BDEW), and automatic import of statutory German public holidays (VDI 4655). The six bundled Bautzen TRY files are selectable from a drop-down; custom TRY files can be loaded via a file browser.

BDEW SigLinDe tab

VDI 4655 tab

Demo scripts

python examples/bdew_demo.py    path/to/TRY.dat --output ./output
python examples/vdi4655_demo.py path/to/TRY.dat --output ./output

The package ships with six TRY files for Bautzen (51.1676°N, 14.4222°E, DWD climate zone 9) in data/try/: average year, extreme cold winter, and extreme hot summer for both the 2015 and 2045 climate epochs.

from pyslpheat import TRY_BAUTZEN_2015, TRY_BAUTZEN_2015_WINTER, TRY_BAUTZEN_2015_SUMMER
from pyslpheat import TRY_BAUTZEN_2045, TRY_BAUTZEN_2045_WINTER, TRY_BAUTZEN_2045_SUMMER

Additional TRY files are available from DWD / BBSR.

Documentation

Full API reference: docs/DOCUMENTATION.md

Related projects & ecosystem

Several Python packages implement BDEW or VDI 4655 heat load profiles. The table below covers the actively maintained, pip/conda-installable tools.

Package	Install	Standards	Subtypes	Hourly shape	Peak scaling	Heating limit	Stochastic	TRY bundled
pyslpheat (this)	pip install .	BDEW SigLinDe, VDI 4655	HEF/HMF 03–05, 33/34; all G-types 01–05, 33/34	BGW hourly factors	✓ Mode B/C (β-bisection)	✓	✓ peak jitter + log-normal + discrete DHW draws	✓ 6 files (2015/2045 ×3)
oemof-demand	pip install oemof-demand	BDEW SLP, VDI 4655	building_class 1–11, wind_class 0/1	BDEW hourly factors	✗	✗	✗	✗
when2heat	pip from repo	BDEW SLP	SFH / MFH / COM only	BGW hourly factors	✗	✗	✗	✗
lpagg	conda	VDI 4655	EFH / MFH	15-min day-type profiles	✗	✗	✓ random time-shift	✗

Broader ecosystem

Tools beyond pip/conda packages — physics-based simulators, activity-based generators, and external web/desktop tools that produce comparable heat demand time series:

Tool	Type	Method	Notes
HiSim	Python package	Physics-based thermal simulation (TABULA building database)	Full dynamic simulation incl. solar gains, internal loads, thermal mass; most detailed approach
pylpg	Python package	Activity-based (LoadProfileGenerator, TU Chemnitz)	Stochastic activity chains per person → energy flows; minute-level resolution
EnSySim	Python package (HS Zittau/Görlitz)	Heating degree-hours (HGT)	Simple linear approach; no BDEW coefficients; useful as lower-complexity baseline
nPro	Web tool (RWTH Aachen)	Proprietary (VDI 4655 & BDEW referenced)	Pre-calculated profiles by building age class; export as CSV
Sophena	Desktop software (IER Stuttgart)	BDEW methodology + regional statistics	District heat planning tool; export profiles as CSV
synPRO	Web tool (Fraunhofer ISE)	Statistical-synthetic 15-min profiles	Stochastically generated; export as .dat; residential old stock / passive house
FreePlan	Excel tool (TU Dresden)	Regression model (model public, training data/coefficients not)	Spreadsheet-based profile generation; no scripting required

What pyslpheat does differently

Original coefficients, then options on top. pyslpheat reads the unmodified BDEW coefficient table (daily_coefficients.csv) covering all residential and commercial profile types with all subtype variants. Configuration options are layered on top — none of them alter the underlying BDEW shape unless explicitly requested:

Option	What it does
subtype	Selects the insulation/consumption class: "03" high (old stock), "04" medium, "05" low (new build); "33"/"34" are the SigLinDe variants with non-zero linear correction terms
dhw_share	Fixes the DHW fraction of the total annual demand, overriding the profile's natural split
dhw_flat	Distributes DHW uniformly across 24 h (temperature- and weekday-independent)
heating_limit_temp	Sets a Heizgrenztemperatur — space heating is zeroed on warm days; total energy is renormalized automatically
heating_exponent	Power-law reshaping of the daily heating distribution: >1 → sharper winter peaks, <1 → flatter (same annual total)
peak_design_kW + design_temperature	Mode B/C — constrains the design-day peak via bisection on a shape exponent β, simultaneously satisfying both the annual energy and the design load target
stochastic=True	Adds day-level peak-time jitter (circular hourly shift) and log-normal amplitude noise with energy renormalization, producing statistically distinct realisations
dhw_draw_events=True	Replaces the smooth DHW baseline with stochastic clustered draw events (bimodal morning/evening, Poisson count per day, log-normal amplitude); annual DHW energy preserved

Relation to DistrictHeatingSim. The BDEW and VDI 4655 modules were originally developed as part of DistrictHeatingSim, an open-source district heating simulation framework (own development). After its publication, the heat profile logic was extracted into this standalone package and further extended — most notably with the stochastic post-processing (peak jitter + log-normal amplitude noise) which was not present in DistrictHeatingSim. pyslpheat can therefore be used independently of the full simulation framework.

demandlib comparison. demandlib.bdew.HeatBuilding parametrises temperature sensitivity via building_class (1–11) and wind_class (0/1) — continuous knobs that shift the sigmoid curve. pyslpheat instead uses the discrete BDEW subtype codes (03/04/05/33/34) that correspond directly to the tabulated coefficient sets in the official guideline, then adds independent post-processing options. Neither approach is strictly superior; they answer different questions.

Know another tool that belongs here? Feel free to open an issue or pull request to add further heat load profile calculation tools to the comparison table.

Development context

pyslpheat emerged from research in heat supply at Hochschule Zittau/Görlitz. The heat profile modules were originally developed as part of DistrictHeatingSim within the SMWK-NEUES TG70 project (development and testing of methods and tools for the conceptualization of sustainable heating networks). They were subsequently extracted into this standalone package and further extended — most notably with stochastic post-processing — as part of an EFRE/ESF-funded Junior Research Group "Energy Storage" at Hochschule Zittau/Görlitz.

Funding

License

MIT — see LICENSE.