Post on 03-Aug-2018
transcript
1 January 30, 2008
Next Generation Low GWP Refrigerant HFO-1234yfPart 2
Honeywell / DuPont Joint Collaboration
January 23, 2008
ASHRAE Meeting, New York, NY
Barbara Minor Mark Spatz
DuPont Honeywell
2 January 30, 2008
Agenda
• Flammability Properties– LFL-UFL
– Minimum ignition energy
– Burning velocity
– Flammability indices
• Flammability Tests and Modeling– Static ignition tests with various sources
– Flame extension test
– CFD modeling
• Risk Assessments
3 January 30, 2008
Refrigerant Flammability Properties
• Is the refrigerant flammable?– LFL – lower flammability limit– UFL – upper flammability limit
• What is the probability of an ignition source being present of sufficient energy to cause an ignition?
– Minimum ignition energy– Autoignition temperature
• What is the impact (damage potential) if an ignition occurs?
– Burning velocity– Heat of combustion
4 January 30, 2008
1234yf Flame Limits (LFL and UFL)
LFL Values
MoreFlammable
Gasoline 1.3 vol.%
Propane 2.2 vol.%
Acetylene 2.5 vol.%
EthyleneOxide 3.0 vol.%
HFC-152a 3.9 vol.%
Methane 4.6 vol.%
1234yf 6.5 vol.%
HFC-32 14.4 vol.%
Ammonia 15 vol.%
ASTM E681 Apparatus
Air In Refrigerant In
Spark Ignition
Stirrer
• ASTM E-681 in US– 2001 version cited by
ASHRAE (12 liter flask, spark ignition)
–Flame must reach the wall and exhibit > 90 degree angle
1234yf Has Narrowest Flammable Region
LFL* UFL* Delta (vol%) (vol%) (vol%)
Propane 2.2 10.0 7.8
R152a 3.9 16.9 13.0
R32 14.4 29.3 14.9
Ammonia 15 .0 28.0 13.0
1234yf 6.5 12.3 5.8* Measured at 21°C
5 January 30, 2008
1234yf Minimum Ignition Energy
• ASTM E-582 apparatus with 1 liter vessel– Tested HFO-1234yf up to 1,000 mJ which
was maximum energy available– No ignition occurred
• “Wall-effects” can quench flame propagation suppressing ignition
– “Wall-effects” have greater significance with difficult to ignite materials (like some halocarbons), especially in smaller vessels
– Important to use appropriately sized test vessels for halocarbon flammability testing
• Retested in 12 liter vessel up to 1,000 mJ – Still no ignition occurred
ASTM E-582 MIE Apparatus• 1 liter spherical vessel• Metal electrodes, variable
gap; vary gap to > ignition quenching distance
• Sight glass to monitor ignition & propagation
> 10001234yf
680Ammonia
0.29Gasoline
>30, <100HFC-32
0.38HFC-152a
0.25Propane
0.47Methane
MIE, mJ
1234yf Is Very Difficult To Ignite . . . Similar To Ammonia
6 January 30, 2008
1234yf Flammability Properties
0.01
0.1
1
10
100
1000
0 1 2 3 4 5 6 7
Static discharge from clothing
Mechanical sparks.Stray current sparks.Ungrounded con-ductors.
Flames.Chemical sources.Propagating brushes.
Lower Flame Limit, vol.%
Min
imum
Igni
tion
Ene
rgy,
mJ
Personnel spark limit
Methane
152a
Acetylene
Propane
Iso-Butane
Gasoline
1234yf
Increasing
Flammability Risk
Ignition Source
A glowing cigaretteWill not ignite methane
(A.D. Little)
7 January 30, 2008
Burning Velocity
ISO 817 Burning Velocity Apparatus
• 18 vol.% in airstoichiometric
• Propagates entire lengthof the tube – 150 cm
• 8 vol.% in air - stoichiometric• Flame propagates < 8 cm
at low velocity • Wall effects extinguish
the flame
150 cm tube
38 mm internal diameter
Ignition Source15 kV, 30 mA power for 0.3 s.
Video area
1234yf Burning Velocity Less Than That Of HFC-32
HFC-32 1234yf
1.5*6.77.22346BV,cm s-1
1234yf32NH3152aPropane
Burning Velocity Data
*3 liter spherical method
8 January 30, 2008
Flammability Indices
( )( ){ } SuQstLFLLFLLFLUFLRF
M
Q
LFL
UFLRF
UFL
LFLF
LFL
CstR
××−×=
×
−
=
−=
=
/2
1
1
Cst = Stoichiometric flammable composition in air, vol.%
Q = Heat of Combustion per one mole
Qst = Heat of Combustion per one mole of the Stoichiometric mixture, kJ/mol
Su = Burning speed in Meters/Second
M = Molecular weight
37.256.70.551.99Propane
17.916.60.51.78152a
2.34.60.331.3132
1.56.80.271.45Ammonia
0.6*3.60.270.971234yf
RF2(kJ/mol)(m/s)
RFkJ/g
FRMolecule Flammability Index1. R proposed by O. Kataoka ,
ISO TC86/SC8/WG5, Feb 2000
2. F proposed by Kondo,J. Hazard. Mater. 2606 (2001) 1-16
3. RF proposed by Kondo, J. Hazard. Mater. A93 (2002) 259-267
4. RF2 proposed by Kondo, Seminar 10, ASHRAE, June 2003.
*Assumes a burning velocity of 1.5 cm s-1
1234yf Has Lowest Flammability Index
9 January 30, 2008
Static Ignition Test Setup
VacuumConnection
Refrigerant/AirConnection
Igniter8.5 Liter
Glass Tube
Rubber Stopper “Overpressure Relief”
PressureTransducer
PressureIndicator
10 January 30, 2008
Static Ignition Tests at Worst Case Conditions
- 8.5 liter chamber with well mixed conditions - Stoichiometric worst case refrigerant/air concentrations - Room temperature, 50% relative humidity for R152a, R32 and HFO-1234yf; dry for ammonia to prevent reaction with water
HFO-1234yf Is More Difficult To Ignite Than R32 and R152a, Similar To Ammonia
Test R152a R-32 Ammonia 1234yf
Cigarette No Ign No Ign No Ign No Ign
Glowing Hot Wire Ign No Ign No Ign No IgnButane Lighter Ign Ign Weak Ign Weak Ign
Fused Wire 100-300 J (worst case) Ign Ign Ign Ign
11 January 30, 2008
1234yf Aerosol Flame Extension Test
ASTM Method D 3065-01
1234yf Is More Difficult To Ignite In Dynamic Situations
• Aerosol flammability test• Referenced by IATA & US DOT• Aerosol (liquid + vapor) spray through a
candle flame from a distance of 15 cm • Flame extension measured
• Flame Extinguished
Vapor
• Flame Extinguished
Mist
• Flame Extinguished
Stream
12 January 30, 2008
CFD Modeling HFO-1234yf vs R-152a
• Used Computational Fluid Dynamics model to simulate a large evaporator leak in a vehicle
– Internal volume = 3.1 m3
– 550 g refrigerant charge
– 6 passengers
– 100% recirculation
– Low fan speed – 60 cfm
• Modeled activation of a squib valve 10 seconds after the leak begins
13 January 30, 2008
1 Second After Leak
16%
14%
12%
10%
8%
6%
4%
12%
11%
10%
9%
8%
7%
152a vol.%Top View Side View
1234yf vol.%
152a
1234yf
UFL
LFL
UFL
LFL
14 January 30, 2008
5 Seconds After Leak
16%
14%
12%
10%
8%
6%
4%
12%
11%
10%
9%
8%
7%
152a vol.%Top View Side View
1234yf vol.%
152a
1234yf
UFL
LFL
UFL
LFL
15 January 30, 2008
9 Seconds After Leak
16%
14%
12%
10%
8%
6%
4%
12%
11%
10%
9%
8%
7%
152a vol.%Top View Side View
1234yf vol.%
152a
1234yf
UFL
LFL
UFL
LFL
16 January 30, 2008
11 Seconds After LeakSquib Activated at 10 Seconds
16%
14%
12%
10%
8%
6%
4%
12%
11%
10%
9%
8%
7%
152a vol.%Top View Side View
1234yf vol.%
152a
1234yf
UFL
LFL
UFL
LFL
17 January 30, 2008
13 Seconds After Leak
16%
14%
12%
10%
8%
6%
4%
12%
11%
10%
9%
8%
7%
152a vol.%Top View Side View
1234yf vol.%
152a
1234yf
UFL
LFL
UFL
LFL
18 January 30, 2008
CFD Modeling Conclusions
• HFO-1234yf flammability envelope was significantly smaller than 152a
• HFO-1234yf flammable region was limited to a small volume at the exit of the vents
• Flammable HFO-1234yf / air compositions did not collect or pool in other regions of the car
• This significantly reduces the possibility that an ignition source with sufficient energy can be present in the flammable region during a leak
• Squib activation eliminated the HFO-1234yf flammable region within one second; 152a flammable region remains
HFO-1234yf Flammability Risk Significantly Reduced Versus 152a
19 January 30, 2008
�For most fires to happen, fuel and air at the right concentration, and an ignition source, with a sufficient energy level must co-exist at the same place and in the same time.
�Several risk assessments based on fault tree analysis are underway in US, Japan and Europe utilizing inputs of modeling and leak experiments
�Release Experiments� Cabin and underhood� Normal operation and crash condition� Service
�CFD modeling to visualize concentration distribution for various scenarios.
R152a
1234yf
Risk Assessments
20 January 30, 2008
HFO-1234yf Summary
• Excellent environmental properties– Very low GWP, Zero ODP, Good LCCP– Atmospheric chemistry determined
• Low toxicity, comparable with 134a– Low acute and chronic toxicity– Significant testing completed
• System performance very similar to 134a– Excellent COP and Capacity, no glide– Thermally stable and compatible with 134a components– Potential for direct substitution of 134a
• Mild flammability (manageable)– Flammability properties significantly better than 152a; – potential for “A2L” ISO 817 classification versus “A2” for 152a– Significantly different vehicle leak behavior than 152a– Potential to use in a direct expansion A/C system; better performance, lower
weight, smaller size than a secondary loop system