Intrinsically Safe Current Limits for Fuel Tanks

1
Intrinsically Safe Current Limits for Fuel Tanks

• Robert Ochs
• FAA/Rutgers University Graduate Fellow

2
Motivation

• Small fragments of cleaning debris, i.e. steel
  wool, when contacted between two charged
  electrodes, can spark, glow and burn
• If this occurred in an environment filled with
  flammable vapors, as in a fuel tank, an explosion
  could occur, resulting in catastrophic damage

3
Definition of Intrinsically Safe

• Any instrument, equipment, or wiring that is
  incapable of releasing sufficient electrical or
  thermal energy under normal operating or
  anticipated failure conditions to cause ignition
  of a specific hazardous atmospheric mixture in
  the most easily ignited concentration ( AC
  25.981-1X)

4
Objective

• To determine the minimum electrical current
  that could cause a flammable fuel/air mixture to
  ignite using a minimum ignition energy test
  apparatus with a calibrated hydrogen-oxygen
  mixture and spark source that can reliably ignite
  the mixture at a determined spark energy

5
Previous Work

• Previous work was done at the FAA Tech. Center
  using an open cup flash point tester to generate
  flammable vapors
• Electrodes with either AC or DC current were
  placed over the cup, and wads of dry and fuel
  soaked steel wool were dropped onto the
  electrodes
• Single filaments of steel wool, aluminum drill
  shavings, and aluminum or brass wool were short
  circuited between the electrodes

6
Previous Test Apparatus
7
Difficulties With Open Cup

• The mixture above the open cup flash point tester
  was not reliably consistent
• Different mixtures require different ignition
  energies
• An apparatus was needed that can deliver
  repeatable mixtures consistently

8
Current Test Apparatus

• Hydrogen-Oxygen-Argon minimum ignition energy
  chamber developed by Lightning Technologies, Inc.
• System can consistently and reliably ignite the
  mixture at 200 microjoules
• Designed to test ignition capabilities of
  electrical failures of system components
• Similar testing with the same materials to be
  tested in this apparatus

9
Hydrogen Chamber
10
Hydrogen-Oxygen Mixture

• Using hydrogen as an ignition detection technique
• Relatively low overpressures
• High probability of ignition at low spark energy
  levels
• Mass flow controllers regulate the amount of each
  gas flowing into chamber

11
Standard Voltage Spark Ignition Source

• High voltage DC power supply
• Variable-vacuum capacitor
• Adjustable spark gap
• Corona source used to initiate breakdown at
  specified voltage

12
Calibrating the Spark

• Set spark gap for 2 mm
• Fill chamber with non-ignitable mixture
• Record breakdown voltage for 10 spark events
• Spark voltage was selected at one std. dev. below
  the mean
• This spark voltage was determined to reliably
  breakdown the gap using the corona source
• The capacitance was calculated for 200 microjoule
  spark energy

13
Calculated Spark Voltage and Capacitance
14
Calibrating the Mixture

• A mixture is desired that will just cause
  ignition at the given spark energy so that the
  minimum energy that will ignite this mixture will
  be no less than 200 microjoules
• Fill tank w/5 VTE to get 99 purity
• Start with 5 H2, 12 O2, and 83 Ar, increase H2
  and decrease Ar until 85-95 ignition probability
  is achieved

15
Ignition Probability

• Total of 21 ignition checks
• 18 ignitions, 3 non ignitions
• 86 ignition probability with this mixture

16
Experimental Setup

• Agilent model 6554A microprocessor-controlled DC
  power supply
• In-line, thin film, non-inductive resistors used
  to dampen transient current overshoot
• Tektronix P5205 voltage probe and Tektronix
  TCP202 current probe, combined with Tektronix
  TDS3014B digital phosphor oscilloscope were used
  to measure and record voltage and current traces

17
Experimental Procedure

• Pre-testing Checklist
• Spark gap checked w/feeler gauge for 2 mm
• Electrodes, glass insulators wiped with isopropyl
  alcohol to avoid current leakage
• Electric space heater used on humid days to
  eliminate water vapor near gap and capacitor
• Mass flow controllers turned on to allow to warm
  up

18
Experimental Procedure

• Verification of ignitable mixture
• Chamber filled with proposed test mixture
• Up to four ignition attempts were made at 200
  microjoules
• If ignition was not achieved, a new chamber fill
  was attempted, if again no ignition, increase H2
  and try again
• When an ignitable mixture is found, testing can
  begin

19
Test Matrix
20
Test Configurations
T.C. 1
T.C. 2
21
Test Configurations
T.C. 3
T.C. 4
22
Test Configurations
T.C. 5
23
Overall Results

• Lowest ignition currents (A)
• Non-inductive resistors used to regulate current
  with 28 VDC
• Steel wool is the only material that caused
  ignition below 100 mA

24
Results Test Config. 1

• Heating of filament was found insufficient to
  cause hot-surface ignition of the gas
• Ignition was only achieved when filament failed
  and burned
• Steel wool has machine oil coating on it,
  inherent in the manufacturing process, may
  initiate or sustain filament burning
• Several brands of steel wool were used in
  testing no significant difference found between
  brands

25
Results Test Config. 1
26
Results Test Config 1

• The length and thickness of a single filament of
  steel wool had a strong effect on the current at
  which the filament would fail and burn
• Non-uniformity of cross sectional area of a
  single filament made it difficult to quantify the
  thickness of the strand
• Measuring the resistance of a filament with an
  ohmmeter gave a good approximation of the
  relative thickness of a filament when compared to
  filaments of the same length
• Filaments of various length and thickness were
  tested to find the lowest failure current

27
(No Transcript)
28
Results Test Config 1

• Several tests were conducted with only air in the
  chamber
• Filament failure/burning was found at currents as
  low as 53 mA in air
• When filaments of similar length/thickness were
  tested with flammable gas, no filament burning or
  ignition was witnessed, only filament failure
• O2 in air 21, O2 in mixture 12
• Reduced O2 concentration in gas mixture may not
  be sufficient for filament burning

29
Results Test Configs. 2 3

• It was found that when a single filament of steel
  wool makes/breaks contact with ground, small
  thermal sparking can be observed
• Thermal sparks were insufficient to ignite the
  mixture
• However, if the filament was left in contact with
  electrode for several seconds, a situation
  similar to test config. 1 was witnessed, and the
  filament would fail and burn

30
Results Test Config 5

• When a wad of steel wool is brought into contact
  with a charged electrode pair, burning of the
  sample was witnessed
• The wad would glow orange and leave behind a
  skeleton of steel wool that still conducts
  current possibly low temperature burning of the
  oil coating
• The burning of the wad was found to ignite the
  flammable mixture at currents as low as 99 mA

31
Results Test Config. 5

• As before, several tests were conducted in air
  only
• It was again found that a wad of steel wool could
  burn with an input current of only 45 mA
• When tested in the flammable mixture, no burning
  of steel wool wads was found below 99 mA
• Again, the oxygen deficiency in the flammable
  mixture inhibits combustion at low currents

32
Conclusions and Recommendations

• The lowest current found to ignite the flammable
  mixture was 99 mA with a wad of steel wool
• Other materials were found to ignite the mixture
  at higher currents, out of the scope of this
  research
• The lowest current that caused failure and
  burning of steel wool was 45 mA in air

33
Conclusions and Recommendations

• It can be assumed that if the same burning
  occurred in air as in the flammable mixture,
  ignition would be achieved
• Future testing should use known calibrated
  flammable mixtures with oxygen concentrations
  similar to that of air