feat:add lake-water-sourced heat pumps

[amos-schledorn]

This PR adds support for lake-water heat pumps, building on the implementation of @MajaHIC and the existing river-water heat implementation.

Method

Lake water heat potential is estimated by:

1. Water temperature: Approximated from ambient air temperature using the empirical formula from Triebs & Tsatsaronis (2022), originally developed for rivers:

   $$T_{water} = k_1 + \frac{k_2}{1 + e^{k_4(k_3 - \bar{T}_{air})}}$$

   where $\bar{T}_{air}$ is a 13-day moving average of ambient temperature.

2. Volume flow: Lake volume is assumed to turn over once per year. The annual volume is converted to an equivalent flow rate in m³/s for consistency with the river-water power calculation:

   $$\dot{V} = \frac{V_{lake}}{8760 \times 3600}$$

3. Thermal power: Calculated using the base class formula:

   $$P = \dot{V} \cdot \rho \cdot c_p \cdot \Delta T$$

   where $\Delta T = \min(\Delta T_{max}, T_{water} - T_{outlet,min})$, with default $\Delta T_{max} = 1$ K.

Data Sources

Data	Source	Notes
Lake polygons & volumes	HydroLAKES v1.0 (Messager et al. 2016)	Vol_total in MCM (million m³)
Ambient temperature	HERA dataset	6-hourly, reprojected to EPSG:3035
Temperature approximation	Triebs & Tsatsaronis (2022), ECOS proceedings	Moving average + logistic formula
ΔT assumption	Triebs dissertation (2023)	1 K/year on lake volume

Changes by File

New files

• scripts/build_surface_water_heat_potentials/approximators/lake_water_heat_approximator.py — Lake heat approximator class
• scripts/build_surface_water_heat_potentials/build_lake_water_heat_potential.py — run script

Modified files

• rules/build_sector.smk — Added build_lake_heat_potential rule
• rules/retrieve.smk — Added retrieve_lake_data rule for HydroLAKES GDB
• scripts/definitions/heat_system.py — Added LAKE_WATER to HeatSource enum and heat pump cost mapping
• config/config.default.yaml — Added lake_water to heat source options
• config/plotting.default.yaml — Added colors for lake water heat components
• config/test/ — Added lake_water as heat source for tests
• doc/release_notes.rst — Updated documentation

Testing

Config

foresight: overnight

scenario:
  clusters:
  - 8
  planning_horizons:
  - 2040

clustering:
  temporal:
    resolution_sector: 25h

countries: ['DE', 'DK']

sector:
  district_heating:
    supply_temperature_approximation:
      max_forward_temperature_baseyear:
        DE: 110
        DK: 80
      min_forward_temperature_baseyear:
        DE: 75
        DK: 55
      return_temperature_baseyear:
        DE: 60
        DK: 50
      relative_annual_temperature_reduction: 0.0
    ptes:
      enable: false
  heat_sources:
    urban central:
    - air
    - lake_water
    - river_water
    
solving:
  options:
    noisy_costs: false

Heat potentials

System costs

Scenario	Objective Value [bn€]	Difference to Air-only [%]
Air-only	117.36	0.00
Air + lakes	117.24	−0.10
Air + lakes + rivers	117.15	−0.18

Heat mix

Air-only

Air + Lakes

Air + Lakes + Rivers

Checklist

• I tested my contribution locally and it works as intended.
• Code and workflow changes are sufficiently documented.
• Changed dependencies are added to pixi.toml (using pixi add <dependency-name>).
• Changes in configuration options are added in config/config.default.yaml.
• Changes in configuration options are documented in doc/configtables/*.csv.
• For new data sources or versions, these instructions <https://pypsa-eur.readthedocs.io/en/latest/data_sources.html>_ have been followed.
• A release note doc/release_notes.rst is added.

[fneum]

Looks good! Some minor comments about new dataset integration.

[amos-schledorn]

Note: 36110d5 fixed a bug that caused the return flow temperature to be used in the COP approximation of all heat sources, including those pre-heating the return flow.
-> The above results assume incorrect COPs for heat sources above the return temperature boosted via heat pumps.

[cpschau]

Just nitpicking about distinction between polyhedron and polygon. Thinking about it a polyhedron is probably nothing else than a set of cohesive polygons, so also fine if it stays as it is.

I was a bit baffled about the assumption about the singular turnover per year for all lakes.

Other than that, this looks good to me. After the changes in #1893, it seems quite straightforward to add new heat sources.