Computational Energy Engineering

We design
CHP Diagnostics
with code.

Inspired by LEAP 71's Computational Engineering Models. 30 years of energy technology expertise, encoded into 19 causal chains and 70 live sensors. A new way to operate modern large-scale CHP plants in the extremely flexible electricity market.

--Score
--Live Sensors
--Active Alarms
19Causal Chains

THE MANIFESTO

Individual problems are seen, but their impact on the overall system isunderestimated.

— Michael Wentzke, IG Biogasmotoren. Wentzke describes the core problem of modern CHP plants: data is available, but not actionable for leadership. Decisions are made based on good feeling, not numbers. We solve this with theDecathlon Logic+ Causal Engine — each of the 10 disciplines measurable, holistically connected, with encoded engineering knowledge.

# Causal Engine — Layer 2: Traversing Causal Chains CAUSAL CHAINS = { "K01": { // Wentzke D3 Gas Mixing "name": "Gas Quality → Engine → Power", "trigger_var": "awite_gasanalyse_rch1_ch4", "trigger_op": "<", "trigger_limit": 50.0, "effect": "Lambda decreases → Cylinder temps rise → Power decreases", "subsystems": [1, 3, 4, 6, 8], }, # ... 18 additional causal chains }

10 DISCIPLINES · DECATHLON LOGIC

Onemeasurable discipline after another.

Methodology after Michael Wentzke (IG Biogasmotoren). Each discipline monitored by our live sensors. Real-time assessment against manufacturer limits and 30 years of field experience.

Loading...

THE COMPUTATIONAL STACK

Knowledgeencoded. Dataconnected. Forecastsvalidated.

LAYER 0
Feeding
Substrate Mix → KTBL → Target Gas Production
LAYER 1
Subsystems
10 Wentzke disciplines, 70 sensors with 3-stage limits
LAYER 2
Causal Chains
19 chains K00-K19, traversing domain knowledge
LAYER 3
Market + Heat
Electricity Price · Weather · 53 District Heating Customers
LAYER 4
Forecasts
Time-shifted Predictions in ClickHouse
LAYER 5
Validation
Learner Confidence per Causal Chain