Relief System Sizing for Runaway Chemical Reactions: A Simple Comprehensive Approach

Thermal Runaway

• A thermal runaway is the progressive production of heat from a chemical process and occurs when the rate of heat production exceeds the rate of heat removal
• The batch temperature rises because there is insufficient cooling available to remove heat from the system to maintain isothermal conditions

Heat Generation > Heat Loss = Thermal Runaway

Hazards Arise from Pressure

• When considering reaction hazards, temperature is rarely a hazard on its own. The impact of temperature rise on the system pressure is much more important.
• There are three potential sources of overpressure:
    -Gas generation from the normal process
    -Vapor pressure effects (as a consequence of heat from the normal process)
     -Heat from the normal process leading to secondary reactions at elevated temperature (causing gas and/or vapor 
      pressure effects)
• Emergency Relief System (ERS) must be designed to safely vent possible sources of overpressure

Upset Scenario Selection

Determine plausible upset scenarios from Process Hazard Analysis (PHA)

What leads to or triggers runaway reaction?

• Incorrect reagents or wrong order of addition
• Reactant accumulation
• Contamination
• Corrosion → unwanted catalytic effects
• Overcharge / undercharge catalyst
• Addition at wrong temperature
• Loss-of-cooling
• Loss-of-mixing
• Inadvertent heating
• Fire exposure
• Material compatibility

Adiabatic Calorimetry

• Low phi-factor calorimetry allows for direct application of data to process scale
• Directly simulate upset scenarios of interest

ARSST Methodology

• Low thermal inertia (phi-factor φ = 1.05)
• Thermal scan to identify moderate to high exothermic activity
• Open system
  - Impose backpressure to suppress boiling
  - Initial pressure depends on goal of test

• Direct measurement of sample temperature 

VSP2 Methodology

• Low thermal inertia (phi-factor φ = 1.05-1.15)
• Simulate normal process or upset conditions
• Identify mild to high exothermic activity
• Open or closed cell
• Uses pressure-balancing technique

Injection Piston

Syringe Pump

Setup for Gas Addition

Source Term Determination for Relief System Design

• System classification
  -Vapor (Tempered)
  -Gassy (Non-tempered)
  -Hybrid, tempered
  -Hybrid, non-tempered
• “Source term” determined based on system classification

Vapor System

• Pressure generation is due to increased vapor pressure
• Latent heat of vaporization (tempering)
• Temperature rise rate is used for vent sizing
• Reaction temperature rise can be controlled by venting

Gassy System

• Generates non-condensable gas
• Latent heat of cooling not available
• Typical of a decomposition reaction yielding gassy products
• Reaction temperature rise cannot be controlled by venting

Hybrid System

• Latent heat of cooling is available at the relief pressure and temperature (tempered)
• Reaction temperature rise can be controlled by venting
• Generates non-condensable gas

Flow Regime Considerations

• Entrained liquid reduces the flow area available for venting

Two-phase flow (foamy)    All gas/vapor flow (non-foamy)

Blowdown Testing in VSP2

• Depressurize VSP2 test cell and determine how much material remains
• Used to mimic superficial velocity to determine expected flow regime

Simple Vent Sizing Formula

• Use with low phi-factor calorimetry data
• Limited material properties required
• Vent size is based on all vapor or gas venting
  -This does NOT mean there is no two-phase flow
  -Two-phase flow can still occur but uncertainties are accommodated by allowing sufficient overpressure above the relief set pressure
• Equation presented here is applicable to critical flow
  • “Relief System Sizing for Runaway Chemical Reactions: A Simple Comprehensive Approach,” 11th Global Congress on Process Safety, 2015, K. Kurko
  • “Vent Sizing Applications for Reactive Systems”, AIChE 2001, 5th Process Plant Safety Symposium, J. Burelbach

A  =  required vent area (m²)

C_D  =  discharge coefficient (-)

m  =  mass of vessel contents (kg)

c_p  =  heat capacity of vessel contents   (J/kg·K)

 =  temperature rise rate from   calorimetry test (K/s)

=  latent heat of vessel contents  (J/kg)

T  =  venting pressure (Pa)

R  =  universal gas constant   (8,314.47 J/kmol·K)

T  =  venting temperature (K)

MW_v  =  molecular weight of vapor   (kg/kmol) 

v  =  freeboard volume of test (m³)

     =  pressure rise rate from   calorimetry test (Pa/s)

m_t  =  mass of test sample (kg)

MW_g  =  molecular weight of gas   (kg/kmol)

Guidelines for Use of Simplified Formula

• Vapor systems
  -Evaluate material properties at set pressure of relief device
  -Use set pressure as venting pressure
  -Use temperature rise rate at set pressure (from calorimetry test)
  -Use of equation requires 40% overpressure (on an absolute basis)
  -For foamy systems, multiply vent area by 2
• Gassy systems
  -Evaluate material properties at peak gas generation rate
  -Use maximum allowable accumulation pressure (1.1×MAWP) as venting pressure
  -Use peak pressure rise rate (from calorimetry test)
• Hybrid systems
  -Evaluate material properties at set pressure of relief device
  -Use set pressure as venting pressure
  -Use temperature and pressure rise rates at set pressure (from calorimetry test)
  
  

Vapor System Example

• 3500 kg of a phenol-formaldehyde resin is produced in a 5 m3 reactor
• Reactor MAWP is 30 psig
• Reaction is run at 50°C
• Sodium hydroxide is added to reaction mixture of phenol, water, and formaldehyde
• Results of PHA indicate fast addition of sodium hydroxide would overwhelm cooling capacity
• Desired set pressure of rupture disk is 10 psig
• Closed test cell VSP2 test run

Vapor System Example (Closed Cell VSP2 Test Data)

Vapor System Example Calculation

Gassy System Example (Open Cell ARSST Test Data, P0 = 88 psig)

• 210 kg of 40% dicumyl peroxide in 2,2,4-trimethyl-1,3-pentanediol diisobutyrate stored in 340 L tank
• Tank MAWP is 80 psig
• Results of PHA indicate a fire in surrounding area would elevate the temperature of the tank to cause decomposition of dicumyl peroxide
• Fire exposure rate determined to be 0.5°C/min
• Desired set pressure of rupture disk is 50 psig
• Open test cell ARSST test run at 88 psig

Gassy System Example Calculation

Hybrid System Example (Open Cell VSP2 Test Data, P0 = 110 psig)

• 2000 kg of 25% hydrogen peroxide is stored in a 700 gallon tank
• Tank MAWP is 100 psig
• Results of PHA indicate iron contamination could cause a runaway reaction due to accelerated decomposition of hydrogen peroxide
• Desired set pressure of rupture disk is 20 psig
• Open test cell VSP2 tests run at 110 psig and 20 psig

Hybrid System Example – Calculation

Summary – Relief Device Sizing

• Determine all credible upset scenarios
• Perform calorimetric tests
  - Use low thermal inertia adiabatic calorimetry
• - Simulate actual upset scenarios
• Apply experimental data to vent sizing formula
  - Minimal physical property data required
  - Results compare well with large scale experimental data

“Without data, all you have is an opinion”