Abstract: This research aims to develop and test a new fire detection system using two levels of temperature sensors. The results show that the proposed system is able to detect fire more accurately and efficiently than traditional systems.

Introduction: Ship fires pose significant safety risks due to difficulties in evacuation and the potential for carrying flammable cargo. Traditional fire detection systems are prone to false alarms, leading to wasted resources and a desensitised response to genuine threats. The placement and sensitivity of sensors often contribute to this issue. This research focuses on developing a more reliable fire detection system using two levels of temperature sensors in a room of Russia’s ship museum Aurora.

Methodology: This research uses a field model to simulate the fire dynamics within a room. The software chosen for this purpose is Fire Dynamics Simulator (FDS), along with its graphical interface PyroSim and the visualisation tool SmokeView. The simulations were run on the “Polytechnic” supercomputer due to the computational demands of the field model. The room design replicates a space within the Aurora cruiser. Eighteen temperature sensors, nine on each of two levels, are strategically placed within the room to gather temperature data during a simulated fire.

Key Findings:

  • Sensors on the upper level consistently recorded rising temperatures with a high standard deviation and rate of change, indicating a clear fire event.
  • Sensors on the lower level exhibited fluctuating temperature readings with a lower standard deviation and rate of change, suggesting temporary heat fluctuations not necessarily indicative of a fire.
  • The sensor closest to the simulated fire source (THCP-H2) recorded a lower standard deviation despite its proximity, attributed to the rapid fluctuations caused by intense heat radiation.
  • The rate of temperature change was higher in upper-level sensors, further supporting the conclusion that fire detection is more reliable with data from both levels.

Significance/Implications:

  • The use of two-level temperature sensors provides a more reliable fire detection criterion. By comparing data from both levels, the system can differentiate between a sustained fire and temporary heat fluctuations.
  • This system can potentially reduce false alarms, improving the efficiency of fire response efforts and preventing desensitisation to real fire events.
  • By analysing the location of the sensor with the highest rate of change, the system can also estimate the fire’s location.
  • Future research could explore integrating the system with neural networks to predict fire events based on sensor data.

Full Text: Link

Citation Information:

Malykhina, G.F., & Shahid, M.N. (2017). Simulation on Supercomputer of Fire Detection System Having Two Levels of Temperature Sensors. Измерения в современном мире-2017, 70-78.