Physics Division

Printer Friendly Copy



Reference Material

Policy and Requirements

Methods of Compliance

Operational Requirements



Properties of Liquids


Relief Valve Sizing


Convective Heat Transfer


Relief Vent Pressure Drops


Oxygen Deficiency Hazards

APPENDIX 3  Oxygen Deficiency Hazards (ODH)

(Taken largely from CEBAF Cryogenic Safety Manual)


 Oxygen Deficiency – the condition of the partial pressure of atmospheric oxygen being less that 135 mmHg (about 18% by volume at a barometric pressure of 740 mmHg at ANL).  [American Conference of Governmental Industrial Hygienists]


 1. A quantitative assessment of the increased risk of  fatality from (potential) exposure to reduced atmospheric oxygen shall be conducted for all operations which are physically capable of exposing individuals to an oxygen deficiency.  This assessment shall specify the Oxygen Deficiency Hazard Class as well as any unusual precautionary requirements.

2.   Precautionary measures shall be implemented according to the ODH Class unless otherwise stated in the risk assessment.  ODH Class 0 is the least hazardous and requires no special precautions.  ODH Class 4 is the most hazardous and requires the greatest precautions.

 Effects of Exposure to Reduced Atmospheric Oxygen

Air normally contains about 21%1 oxygen with the remainder consisting mostly of nitrogen.  Individuals exposed to reduced-oxygen atmospheres may suffer a variety of harmless effects.  Table I contains a list of some of these effects and the sea level oxygen concentrations at which they occur.  At higher altitudes the same effects generally occur at greater volume concentrations since the partial pressure of oxygen is less.  If exposure to reduced oxygen is terminated early enough, effects are generally reversible.  If not, permanent central nervous system damage or lethality result.  Major effects hindering escape from the vicinity of an oxygen deficiency are disorientation and unconsciousness.


In general, the intensities of the effects increase rapidly with falling oxygen concentration and longer exposure duration:  reduced abilities, then unconsciousness, then death.  It can be concluded that any exposure to an atmosphere containing less than 17% oxygen presents a risk.

1Although this section is written in terms of %O2 at sea level, the preferred index of hazard is partial pressure of O2.  Percent O2 is used here to maintain consistency with the “readouts” on oxygen monitors.


Effect Thresholds for Exposure to Reduced Oxygen

(Healthy Individuals at Sea Level) 

Volume % Oxygen





Night vision reduced

Increased breathing volume

Accelerated heartbeat




Reaction time for novel tasks doubled



Impaired attention

Impaired judgment

Impaired coordination

Intermittent breathing

Rapid fatigue

Loss of muscle control



Very faulty judgment

Very poor muscular coordination

Loss of consciousness

Permanent brain damage



Inability to move





Spasmatic breathing

Convulsive movements

Death in 5-8 minutes


ODH Risk Assessment

The goal of ODH risk assessment is to estimate the rate at which fatalities will occur as a result of exposure to reduced-oxygen atmospheres.

 Since the level of risk is tied to the nature of the operation, the fatality rate shall be determined on an operation-by-operation basis.  For the given operation several events may cause an oxygen deficiency.  Each even has an expected rate of occurrence and each occurrence has an expected probability of killing someone.  The oxygen deficiency hazard fatality is defined as:


where   f    = the ODH fatality rate (per hour)

            Pi  = the expected rate of the ith event (per hour), and

            Fi  = the fatality factor for the ith event.

The summation shall be taken over all events which may cause oxygen deficiency and result in fatality.  When possible, the value of Pi shall be determined by operating experience at ANL; otherwise, data from similar systems elsewhere or other relevant values shall be used.

Estimates of “spontaneous” equipment failures rates are given in Tables II and III.  The former contains median estimates collected from past ODH risk assessments at Fermilab.  The latter contains values derived from the nuclear power industry.

General human error rate estimates are presented in Table IV.  Table V lists conservative estimates of the rate of human error as a function of task type and time limit.


Fermilab Equipment Failure Rate Estimates

Failure Mode

Estimated Media Failure Rate






Leak or Rupture

3 x 10-5/hr



Leak or Rupture

1 x 10-6/hr


Electrical Power Failure


Time Rate

(Time Off)

1 x 10-4/hr

(1 hr)




Fluid Line (Cryogenic)

Leak or Rupture

3 x 10-6/hr


Magnet (Cryogenic)

Leak or Rupture

1 x 10-6/hr


U-Tube Change



Small Event

Large Event

1 x 10-3/hr

4 x 10-5/hr



Human Error Rate Estimates

Estimated Error                                                                                   Activity
Rate (D-1)

10-3                  Selection of a switch (or pair of switches) dissimilar in shape or location to the desired switch (or pair of switches), assuming no decision error.  For example, operator actuates large-handled switch rather than small switch.

3 x 10-3            General human error of commission, e.g., misreading label and therefore selecting wrong switch.

10-2                  General human error of omission where there is no display in the control room of the status of the item omitted, e.g. failure to return manually-operated test valve to proper configuration after maintenance.

3 x 10-3            Errors of omission, where the items being omitted are embedded in a procedure rather than at the end as above.

1/x                   Given that an operator is reaching for an incorrect switch (or pair of switches), he selects a particular similar-appearing switch (or pair of switches), where x = the number of incorrect switches (or pair of switches) adjacent to the desired switch (or pair of switches).  The 1/x applies up to 5 or 6 items.  After that point the error rate would be lower because the operator would take more time to search.  With up to 5 or 6 items he doesn’t expect to be wrong and, therefore, is more likely to do less deliberate searching.

10-1                  Monitor or inspector fails to recognize initial error by operator.  Note:  With continuing feedback of the error on the annunciator panel, this high error rate would not apply.

10-1                  Personnel on different work shifts fail to check condition of hardware unless required by check or written directive.

5 x 10-1            Monitor fails to detect undesired position of valves, etc. during general walk-around inspections, assuming no checklist is used.

.2 - .3              General error rate given very high stress levels where dangerous activities are occurring rapidly.

2(n-1)x               Given severe time stress, as in trying to compensate for an error made in an emergency situation, the initial error rate, x, for an activity doubles for each attempt, n, after a previous incorrect attempt, until the limiting condition of an error rate of 1.0 is reached or until time runs out.  This limiting condition corresponds to an individual’s becoming completely disorganized or ineffective.



Human Error Rate as a Function of Response Time

Response Time(s)

Maximum Estimated Error Rate (D-1)

Skill Based Task

Rule Based Task

Knowledge Based Task





















5 x 10-1





Skill-Based Task – An individual initiates a single-step learned response upon receipt of an unambiguous sensor cue.  (Example:  A lone worker initiates escape upon hearing an oxygen deficiency alarm.)

Rule-Based Task – An individual or small group of individuals diagnoses and initiates corrective actions for a simple problem given limited or ambiguous input.  (Example:  Several workers decide whether or not to escape given that one of them passes out but no oxygen deficiency alarms sound.)

Knowledge-Based Task – A group of individuals diagnoses and initiates corrective actions for a novel and/or complex problem.

The value of Fi is the probability that a person will die if the ith event occurs.  This value depends on the oxygen concentration, the duration of exposure and the difficulty of escape.  For convenience of calculation, a relationship between the value of Fi and the lowest attainable oxygen concentration is defined (Figure 1).  The lowest concentration is used rather than an average since the minimum value is conservative and not enough is understood to allow the definition of an averaging period.  If the lowest oxygen concentration is greater than 18%, then the value of Fi is zero.  That is, all exposures above 18% are defined to be “safe” and to not contribute to fatality.  It is assumed that all exposures to 18% oxygen or lower do contribute to fatality and the value of Fi is designed to reflect this dependence.  If the lowest attainable oxygen concentration is 18%, then the value of Fi is 10-7.  This value would cause 0 to be 10-7 per hour if the expected rate of occurrence of the event were one per hour.  At decreasing concentrations the value of Fi should increase until, at some point, the probability of dying becomes unity.  That point was selected to be 8.8% oxygen, the concentration at which one minute of consciousness is expected.

Fig. 1. Graph of the logarithm of the fatality factor (Fi) versus the lowest attainable oxygen concentration which can result from a given event.  This relationship should be used when no better estimate of the probability of fatality from a given event is available.

The oxygen concentration is a confined volume during and after a release of inert gas may be approximated from the following differential equation




            V = the confined volume (ft3 or m3)

            C = the concentration of oxygen

            R = the spill rate into the confined volume (cfm or m3/s)

            Q = the rate of ventilation through the confined volume (cfm or m3/s).

In order to solve this differential, the following assumptions are made:

·           Complete, instantaneous mixing takes place in the confined volume

·             V, R, Q, and the total pressure remain constant

·             The initial oxygen concentration is 21%.

Therefore, the oxygen concentration during the release is


where t is the time from the start of the release.  After the release has ended, the oxygen concentration is



where t is the time after the end of the release (when R becomes zero) and te is the duration of the release.


Once the ODH fatality rate (f) has been determined, the operation shall be assigned an ODH class according to Table VI.


Oxygen Deficiency Hazard Class

ODH Class

f (hr-1)




< 10-7


≥ 10-7 but < 10-5


≥ 10-5 but < 10-3


≥ 10-3 but < 10-1


≥ 10-1

ODH controls

Protective measures shall be implemented in a fashion which reduces the excess risk of fatality from exposure to an oxygen deficient atmosphere to no more than 10-7 per hour.  The following logic tree describes suggested minimum control measures to allow an individual to participate in a given ODH Class 1 or greater operations.


ODH Control Measures


ODH Hazard Class

Environmental Controls






  1.  Warning Signs





  2.  Installed Oxygen Monitor





  3.  Ventilation






ODH Qualified Personnel Controls



  4.  Medical Approval as ODH Qualified





  5.  ODH Training





  6.  Personal Oxygen Monitor





  7.  Self-Rescue Supplied Atmosphere Respirator





  8.  Multiple Personnel in Communication





  9.  Unexposed Observer





10.  Self-contained Breathing Aparatus






ODH Restricted Personnel Controls



11.  Must not be ODH Excluded





12.  ODH Briefing





13.  Self-Rescue Supplied Atmosphere Respirator





14.  One-to-One Escort by ODH Qualified Personnel





X = Required
N/A = Not applicable, ODH restricted personnel shall not be exposed to ODH Class 4 operations

An ODH risk assessment should include a discussion of each of the following:


1.                  Significant potential of reduced oxygen.

2.                Mechanisms

a.      Spontaneous failures

b.         Personnel-mediated failures

i.             Operator error

ii.            Accidents

3.                Operations

a.         Steady state

b.           Other

i.             Start-up

ii.      Repairs

iii.          Special operations

iv.          Shutdown

4.                  Gas dynamics

a.           Ventilation

i.             Natural

ii.            Forced

b.             Stratification/mixing

c.             Diffusion

5.               The bases used for conclusion.

6.                Special requirements

a.                   Area oxygen monitors

b.                  Self-rescue supplied atmosphere respirators

c.                   Unusual procedures

Example:  Oxygen-Deficiency Hazard Analysis in BGO Area

by L. M. Bollinger

The hazard of interest is the liquid nitrogen in a standard 160-l dewar.  This dewar is located in a partially enclosed space 12 ft. high, 34 ft. long, and ~14 ft. wide, with a 3.5-ft. wide opening at each end.  The opening extends from floor to ceiling, and at the top of each opening a fan with a capacity of ~1200 CFM blowing out of the room, thus forcing air to circulate into the room at floor level through each opening.  See layout of area in attached sketch.

From the above data one calculates that the volume of the room is ~160 m3 and the forced ventilation capacity if ~1.1 m3/sec.  In addition, because both spacings are from floor to ceiling, the inherent ventilation caused by convection and diffusion is good.

Analysis – use Fermilab approach in Document #5064

Scenario #1

The most probable accident is that the LN2 line between the dewar and the distribution manifold breaks.  Janssens estimates that the breakage probability is ~10-1 yr-1.  Even if the LN2 line is severed at the output from the dewar, the LN2 emission rate is only ~0.3 l/sec, as measured on 3/1/91 at a dewar pressure of 23 psi.  At this pressure, after vaporizing and warming to room temperature, the expansion ratio for LN2 is [(1.3054}1.134 x 10-3]-1 = 675.  Thus the N2 gas emission rate is 0.20 m3/sec.

Assuming complete mixing for the N2 and air (a reasonable assumption for such good ventilation and low N2 emission) the minimum concentration of O2 in room is

where   Q º exhaust rate = 1.1 m3/sec

            R º emission rate = 0.2 m3/sec

            V º volume of N2 = 0.160 (675) = 108 m3

            Vo º volume of room = 160 m3.


                                                                                            = 0.178

The partial pressure corresponding to this is


According to the Fermilab safety document 5064, the probability of fatality F is




Thus, the fatality rate for scenario #1 is

                                                     f = 10-1 (2 x 10-7) = 2 x 10-8 yr-1,

which is small enough to be ignored.

Scenario #2

The worst credible incident for the LN2 dewar is one in which the main LN2 tube at the top of the dewar is accidentally severed.  This would require a massive blow since the tube is solid SS.  This scenario is similar to scenario #1 except that there the line is restricted to a 3/8” orifice by the output fitting whereas here the 5/8” dewar line is 5/8” dia.  Since the line itself plays a significant role in limiting the flow of LN2 in both cases and since one is dealing with 2-phase flow in both cases, it is estimated that the flow rate of scenario #2 relative to scenario #1 cannot be increased more than the ratio of output-aperture areas.  That is, the flow rate is <0.3 (5/3)2 = 0.83 l/sec.  At this flow rate,


                    =  0.139




            F = 10-3.8 = 1.6 x 10-4.

The probability of such an accident can be estimated from experience in the Physics Division, where no such accident has ever occurred in about 40 years of experience.  At least 10 dewars are present in the building on an average.  Thus, the accident rate P is

                                                P (400 yr)-1 = 2.5 x 10-3 yr-1

The facility rate for this scenario is

               f < (2.5 x 10-3) (1.6 x 10-4) = 4 x 10-7 yr-1

Again, this risk is small enough to be ignored.

Scenario #3

Consider the consequences if all of the LN2 content of the dewar were released very rapidly.  This scenario is not credible since no large source of energy is present in the area.

To be specific, let the LN2 be emitted and be vaporized at a rate of 5 l/sec.  This gives 3.25 m3 of gas per sec.  In this case,


           + 0.117,

For this value,

                                                      F = 10(6.5-8.8) = 5 x 10-3.

Thus, using same P as in scenario #2

f (2.5 x 10-3) (5 x 10-3) = 1.25 x 10-5 y-1 = 1.4 x 10-9 hr-1

This value is small enough that no special precautions are needed.  Moreover, the actual fatality rate is much smaller than f because the occupancy rate of the area is very small (~0.02) averaged over a year.


The LN2 dewar in the BGO area is not a significant ODH hazard.