AI Platform
for Combustion
Get A Free Demo
WHY USE PYRETEK?
Lower Energy Consumption
Higher Production
Safety Improvement
Lower Downtime
Lower Emission
TYPES OF DATA ANALYSED
Operational Data
Furnace Temperature, Furnace Pressure, Burner Power
Energy Consumption Data
Burner Power, Fuel/O2/Air Flowrates.
Safety And Reliability Data
Shutoff valves, Pressure and Temperature switches, UV sensors, Pressure and Temperature sensors
Emission Data
NOx/O2/CO sensors
Lower Energy Consumption
Higher Production
Safety Improvement
Lower Downtime
Lower Emission
HOW DOES IT WORK?
Lower Energy Consumption
Ensuring Safety Device Reliability And Detecting Operational Or Maintenance Errors For Secure, Seamless Operations
Higher Production
Driving CO₂ And NOx Reduction Through Optimized Energy And Oxygen Consumption Efficient Burner Management, And Minimized Unplanned Shutdowns.
Safety Improvement
Proactively Predicting Device Failures And Ensuring Compliance With Emission Permits For Sustainable Operations
Lower Downtime
Maximizing Metal Yield, Minimising Downtime, And Eliminating Bottlenecks For Streamlined, High-Efficiency Production
Lower Emission
Enhancing Precision And Stability in Flow Control, Metering, And Pressure Measurement For Optimal Performance
THEY TRUST US
Use Cases
Explore how Pyretek optimizes combustion processes across various industrial applications.
Burner Flow Accuracy Analysis
Goal
Identify any issues with Natural Gas, Oxygen, and Air flows to the burner for proactive maintenance
Context
Theoretically, flows through a gas train depend only on the Flow Control Valve (FCV) opening and incoming gas pressure
Analysis
- Our AI model uses historical data to statistically predict actual burner flows (PV) using FCV and Pressure sensor readings with 3% accuracy (see chart below)
Outcomes
- If the predicted PV is more than 5% off compared to the actual PV reading, then a proactive warning will be sent to the relevant plant personnel
- Maintenance teams can then check flowmeter calibrations as well as any issues with the flow control valves or pressure sensors
Flow Control Loop Performance Analysis
Goal
Monitor and detect any issues with the burner flow control system for energy efficiency
Context
Combustion systems are considered efficient when actual burner flows (Process Values or PV) closely follow the flow set points (SP) entered by the operators
Analysis
- Overall, NG and Oxygen flow loops are extremely efficient; response times were less than a second and the average difference between SP and PV was less than 1%
- Air flow loops also showed excellent response times
- The difference between SP and PV, however, was found to be significantly higher
- We recommend servicing the combustion air flow control system ASAP
Outcomes
- Anomalies in flow control loop performance can be detected in real-time to send timely notifications for prompt fixes and maintain optimal performance
- AI based recommendation for proactive / predictive servicing based on historic patterns
Reverb Temperature Analysis
Goal
Examine the factors that have the strongest effect on the reverb furnace temperature
Context
Since the reverb furnace temperature is one of the most critical factors affecting furnace productivity, its prediction can help alleviate potentially detrimental operating conditions
Analysis
- Our AI models leverage complex relationships across 30+ variables to predict the target variable of choice and were able to predict Reverb Temps within 5% accuracy
Outcomes
- If the predicted reverb temperature is more than 10% off compared to the actual reading, then a warning will be sent to the relevant plant personnel
- Maintenance teams can then check for temperature measurement issues as well as any issues with the furnace operational parameters
NOx Measurement Analysis
Goal
Examine the factors that have the strongest effect on the furnace NOx generation
Context
Due to stringent NOx regulations, furnace NOx generation severely affects furnace productivity, therefore we believe that our ability to predict it can help alleviate operating conditions that affect NOx adversely
Analysis
- Similar to reverb temperatures, our AI models are able to predict furnace NOx generation within 5% accuracy
Outcomes
- Our model will help identify operating conditions that minimize NOx generation while maintaining optimum production levels
FREQUENTLY ASKED QUESTIONS
Have a question? We are here to help