Build a CCGT Power Efficiency Optimizer with CrewAI
Optimizing heat-rate performance for Doosan Enerbility's $670m combined-cycle gas turbine (CCGT) project in Oman requires multi-agent thermal analysis. Build a multi-agent system using CrewAI to analyze steam cycle efficiency, ambient heat impacts, and fuel intake ratios.
What you are building
The core problem, expected build, and operating context for this challenge.
Utilize CrewAI agent roles to analyze CCGT thermodynamic cycles and optimize power output under Middle Eastern ambient temperature conditions.
How work is evaluated
Evaluates power output optimization calculation accuracy under high ambient temperature conditions.
Shared data for this challenge
Review public datasets and any private uploads tied to your build.
How submissions are scored
These dimensions define what the evaluator checks and which criteria separate a passable run from a strong one.
cooling_active_check
Ensure inlet air cooling activates when ambient temp > 40°C
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
heat_rate_accuracy
Accuracy of calculated efficiency percentage • target: 0.97 • range: 0.9-1
This dimension contributes its full weight only when the submission satisfies the requirement. Partial credit is not awarded.
What you should walk away with
Construct CrewAI agent teams: Thermal Engineer, Fuel Optimization Lead, and Maintenance Specialist
Model gas turbine efficiency degradation at peak ambient desert temperatures (45°C+)
Automate fuel-to-air ratio adjustments to optimize heat rate and reduce CO2 emissions
Generate daily plant performance optimization summaries for plant operators
Reference links and supporting material
Telemetry logs, gas turbine performance curves, and weather data from Oman combined-cycle power facilities.
How this agent runs
Evaluates power output optimization calculation accuracy under high ambient temperature conditions.
Challenge input
JSON plant ambient conditions
CrewAI
Multi-agent framework ideal for collaborative engineering task decomposition
Evaluated output
JSON turbine control profile
- Ensure inlet air cooling activates when ambient temp > 40°C
- Accuracy of calculated efficiency percentage • target: 0.97 • range: 0.9-1
- Benchmark: ThermalPowerBench
- Heat Rate Accuracy target: 97%
- 1 public reference case
- Python execution harness
- Python sandbox (unavailable on Versalist)
View technical recipe
Configured tools
- CrewAI · Required
- crewAI · Optional
- AutoKeras · Optional
Evaluation contract
- cooling_active_check · Weight 1
- heat_rate_accuracy · Weight 1
Recipe state
This is a preview. The configuration can change before the evaluation recipe is locked.