Skip to content
AstemLab

Investor overview

Transforming industrial and urban waste heat into low-carbon energy

The full case, in sequence: the paradox, the concept, the measured opportunity, the technology, the roadmap and the numbers.

Section 01

The industrial and urban heat paradox

Refineries and petrochemical plants discharge low-grade heat continuously through cooling towers and process effluent. Metro tunnels add dense passenger heat and stay warm year-round. All of it is vented to the atmosphere — while boilers a few hundred metres away burn natural gas to heat feedwater.

722,544kW

Gross waste heat

18–30°C

Source temperature

82

Bought and burned as gas

27

Select a station

Section 02

Technology & AI engine

Capture a stable low-grade source, upgrade it with heat pumps under machine-learning control, deliver usable heat. Two delivery paths share one architecture.

1840 °CCaptureCOP 3.8+Upgrade1 + 2.8 = 3.8 kWh6090 °CIndustrial supply
  1. 01

    Capture

    Extract waste heat from refinery liquid effluent, process cooling loops and cooling towers. Sensors monitor flow rate and temperature continuously.

  2. 02

    Upgrade

    AI-governed water-to-water and air-to-water heat pumps amplify temperature using minimal electricity. Machine learning ingests real-time flow, temperature and steam demand to optimise compressor speeds — without disrupting primary operations.

  3. 03

    Industrial supply

    Deliver pre-heated boiler feedwater directly to refinery boilers, cutting natural gas combustion.

  4. 04

    Facility supply

    Deliver space heating and domestic hot water to administrative complexes and district heating networks.

Section 03

Measured, not estimated

27 Baku Metro stations and 98 exits surveyed. Every figure on this site is reproducible from that dataset and four openly published constants.

541,908kW

Recoverable

2,373,557MWh

Annual energy recovered

474,711t

CO₂ avoided annually

237,356

Equivalent homes heated

AssetsFocus
Heydar Aliyev Oil RefineryCooling towers & effluent loops
SOCAR Polymer & CarbamideProcess steam vents & secondary cooling
Sumgayit Chemical Industrial ParkClustered manufacturing waste-heat nodes
SOCAR TowerAdministrative complex heat & cooling demand
Section 04

Why Baku beats the precedents

Higher passenger density means more waste heat per node, which means better economics on the same hardware. London Bunhill 2 and Warsaw M2 prove the concept works abroad; density is the argument that it works better here.

8.3

Baku — million passengers per station

3.7

London — million passengers per station

3.8+

Industrial

3.0

Urban / metro

The concept is proven abroad. The AI control layer and the Baku measurement set are what is new here. London Bunhill 2 (Underground heat recovery, 500+ homes heated) · Warsaw M2 (Metro heat-recovery system)

Section 05

Commercial roadmap

Phase 1

Internal optimisation

Reduces tunnel temperatures and extends equipment lifespan without relying solely on the existing ventilator fleet.

Phase 2

External commercialisation

Sells upgraded hot water below the cost of grid electricity or gas heating.

Phase 3

National integration

Direct alignment with Azerbaijan's 2030 Green Energy Goals, unlocking subsidies and global offset markets.

Section 06

Pilot feasibility

1.8MW

Thermal capacity

11.9GWh

Annual energy saved

476,000AZN

Annual cost savings

4.9–10years

Payback period

Payback differs by case and is shown separately. The two studies model different capital costs against the same module.

The ask

Fund the Phase 1 pilot. Validate the AI digital twin. Scale across the metro network and the petrochemical portfolio.

Figures on this site derive from the 2025 Baku Metro measurement programme.

Request Technical Audit