(104g) Optimal Layout of Gas Detector Networks: A Comparison Study | AIChE

(104g) Optimal Layout of Gas Detector Networks: A Comparison Study

Authors 

Benavides-Serrano, A. - Presenter, Texas A&M University
Laird, C. D. - Presenter, Texas A&M University
Legg, S. - Presenter, Texas A&M University
Mannan, D. M. S. - Presenter, Mary Kay O'Connor Process Safety Center

Optimal layout of gas detector networks: A comparison study.

 Alberto Benavides-Serrano1, Sean Legg2, Carl Laird2 and M. Sam Mannan1,

 (1)Mary Kay O'Connor Process Safety Center, Artie McFerrin Department of Chemical Engineering, Texas A&M University, College Station, TX, (2) Artie McFerrin Department of Chemical Engineering, Texas A&M University , College Station, TX

Gas Detection Systems (GDS) constitute a suitable and widely used Layer Of Protection (LOP) that provides alarms and initiates mitigation for confirmed process leaks.  In order for this LOP to effectively serve its intended purpose an appropriate layout of the gas detector networks must be guaranteed. Currently, both industry and regulators entrust the task of gas detector placement to prescriptive/qualitative processes strongly based on rules of thumb. Results are constrained by the depth that experts and guidelines are able to provide.

Existing prescriptive approaches can result in sub-optimal answers to the layout problem. To find an optimal layout one seeks to maximize the potential for detection success in the event of release while taking into account variables such as process plant and equipment, detection equipment, substance properties, environmental factors, safety issues and cost. We have developed a mathematical programming formulation to determine an optimal layout for gas detectors within actual process facilities. [Legg et al[1]]. This approach is a multi-scenario, mixed-integer linear programming (MILP) formulation considering a large number of potential leak scenarios. Here, we present the details of this technique, and compare the effectiveness of this approach with a number of existing qualitative techniques. The study was conducted making use of a rigorous CFD dispersion study with actual geometry from a process facility. Results provide evidence of the need for furthering the development of systematic gas detector network layout optimization approaches, based on CFD simulations and mathematical optimization concepts.


[1] Legg, S., Siirola, J., Watson, J.P., Davis, S., Bratteteig, A., Laird, C. (2011). A Stochastic Programming Approach for Gas Detector Placement in Process Facilities. Accepted by FOCAPO 2011.