Professional Documents
Culture Documents
WE n7071236 v1 Z Bozic & S Hughes Presentation Some Contentious Issues in The Earth Grid Design in Aust
WE n7071236 v1 Z Bozic & S Hughes Presentation Some Contentious Issues in The Earth Grid Design in Aust
Presented by:
Dr Zoran Božić & Mr Steve Hughes
ABOUT THE PRESENTATION
The objective is to:
• Provide an overview of contentious issues
• Serve as a catalyst to resolve the issues
It is organized in two parts:
• Top down view: broader perspective and
considerations
• Bottom-up view: critical review of individual
building blocks
we_n7071236_v1.ppt 2
ACKNOWLEDGEMENT
The authors would like to acknowledge the
contribution given by Dr Hoohman
Dehbonei and Mr Sandeep Magan and for
the stimulation discussion
Dr Dehbonei and Mr Magan are members
of the ENA EG-0 working group
we_n7071236_v1.ppt 3
BACKGROUND (1)
July 2009, a new issue of AS 2067-2008:
Substations and High Voltage Installations
exceeding 1kV a.c.
Aug 2009, draft EG-0 Guide on Power
System Earthing – Part 1: Management
Principles
Feb 2010, 2nd draft EG-0
we_n7071236_v1.ppt 4
BACKGROUND (2)
Standing obligation:
• Under the Western Australian (Supply and
System Safety) Regulations 2001, Western
Power and other licensed power system
network operators in WA have an obligation to
comply with the requirements laid down in
Australian Standard AS 2067.
we_n7071236_v1.ppt 5
CONTENTIOUS ISSUES (1)
Change in the Australian approach and
inconsistency between AS2067 and EG-0
• Probability of fatality calculation
• Permissible voltage thresholds – shall they exist or
not
Application of the IEC60479 I vs t curves
• Pfibrillation and AS2067
• Pfibrillation and EG-0
Aggregate risk to business
Risk assessment data and role for regular
auditing
we_n7071236_v1.ppt 6
CONTENTIOUS ISSUES (2)
Application of the probabilistic concepts
• Probability of coincidence calculation
• Probability of fatality calculation
• Assessment of risk due to step voltages
• Electric shock duration:
credible contingencies and autoreclosing
• Table of permissible risk limits
we_n7071236_v1.ppt 7
PART 1
by
Steve Hughes
we_n7071236_v1.ppt 8
Change in the Australian approach and
inconsistency between AS2067 & EG-0 (1)
In the soon to be issued ENA EG-0 Power System
Earthing Guide the probability of a fatality resulting from
the heart of a person entering ventricular fibrillation due to
contact with an external transient voltage is stated to be:
Pfatality = Pcoincidence x Pfibrillation
we_n7071236_v1.ppt 10
Change in the Australian approach and
inconsistency between AS2067 & EG-0 (3)
No disagreement with this line of reasoning if the
total level of exposure was restricted to just one
potential hazard.
However this will not be the case in practice and
a responsible organisation will want a clear
understanding and assessment of its total level
risk exposure.
Consequently each study necessitates an
assessment of Pfibrillation , irrespective of the value
determined for Pcoincidence, so all relevant
information is available for risk assessment and
auditing needs etc.
we_n7071236_v1.ppt 11
APPLICATION OF THE IEC60479
I vs t (I/t) CURVES (IEC)
Pfibrillation and AS2067
Pfibrillation and EG-0
we_n7071236_v1.ppt 12
IEC - Pfibrillation and AS2067 (1)
AS 2067 – 2008 states:
‘As a guide, the representative touch voltage (Vt)
limits that may be applied to accessible metal
work for the two typical cases are shown in
Appendix A (Figure A2 and Table A1). These
two cases are described as follows:
C1 = Special locations: High public contact
likelihood assuming bare hands and feet and
thus negligible series impedance
C2 = Normal locations: Normal public contact
likelihood with typical mix footwear and thus
additional series impedance
we_n7071236_v1.ppt 13
IEC - Pfibrillation and AS2067 (2)
we_n7071236_v1.ppt 14
IEC - Pfibrillation and AS2067 (3)
The material presented in AS2067-2008 raises
ambiguity as to exactly how the two curves
presented in AS2067 Figure A2 were derived?
Were they produced by selecting IEC60479
curve c1 (or c2) and using two different values
for an equivalent human body impedance, or,
were they based on the two separate curves (i.e.
c1 and c2) bounding the AC-4.1 zone
(Probability of ventricular fibrillation increasing
up to about 5%) and a single equivalent human
body impedance (of approximately 700 ohms)?
we_n7071236_v1.ppt 15
IEC - Pfibrillation and AS2067 (4)
The correct answer is believed to be by using
IEC/TS 60479-1 curve c1 and two different
equivalent series impedances, which would imply
AS2067 – 2008 requires that the body current
versus time curve not exceed the limits defined
by curve c1.
This would result in Pfibrillation being equal to 0 and,
based on the current safety policy in WA,
precludes using the range of Pfibrillation being
suggested by EG-0.
we_n7071236_v1.ppt 16
IEC - Pfibrillation and AS2067 (5)
One possible way of reconciling the two
documents is for AS2067 to reference
limits based on the body current versus
the duration of current flow, as described
in IEC/TS 60479-1, rather than Vt curves.
we_n7071236_v1.ppt 17
IEC - Pfibrillation and AS2067 (6)
we_n7071236_v1.ppt 18
IEC - Pfibrillation and EG-0 (1)
In appendix B of EG-0 the probability assessment of
Pfibrillation is based on the body current curves contained in
IEC 60479-1 and a much broader range of equivalent
touch and step impendence, which in turn, produce a
wider range of representative touch voltage (Vt) limits to
that presented in AS2067-2008.
A wider difference is expected to pose acceptable risk
when a business can impose strict access and personal
protection equipment requirements but greater
consideration needs to be given to situations that involve
the involuntary exposure of people to direct or
transferred voltage transients.
we_n7071236_v1.ppt 19
IEC - Pfibrillation and EG-0 (2)
One also needs to consider whether the
approach taken by EG-0 to the type and
percentage use of footwear for societal
risk – which is a principle determinant
affecting the magnitude of body current
flow – is a reasonable one?
we_n7071236_v1.ppt 20
IEC - Pfibrillation and EG-0 (3)
While it is generally agreed the likelihood of a
fatality occurring due to the involuntary exposure
to direct or transfer voltage transients is
extremely small, any organisation with a
significant number of assets sited in the public
domain will invariably find that its level of overall
risk will likely fall in a ‘very low’ rather than
‘negligible’ category.
Hence, those accountable for the duty of care of
the business need to seriously considered just
how pliable they want this limit to be?
we_n7071236_v1.ppt 21
IEC - Pfibrillation and EG-0 (4)
While under EG-0 the assessment of risk is
based on probabilistic principles the
understanding and quantifying of the
deterministic values for such parameters as:
the value of earth grid impedance; the fault
current magnitude; the fault clearance time; the
rise of earth potential, etc; still applies.
As described in IEEE Std. 80, for example, these
deterministic parameters will form the basis for
where to start in assessing and quantifying
Pfibrillation
we_n7071236_v1.ppt 22
IEC - Pfibrillation and EG-0 (5)
The determination of the magnitude of these parameters
is general practice for system and plant designers and
the results can be used to produce anticipated worse
case I/t curve for the situation(s) under study.
By superimposing the resulting current/time zone curves
on those presented in IEC 60479 the probability for
Pfibrillation. can be clearly assessed and decisions made,
using cost benefit analysis methods etc, to determine
what, if any, design changes or remedial measures
should be taken.
The benefit of this approach is the values presented for
Pfibrillation are always transparent.
we_n7071236_v1.ppt 23
IEC - Pfibrillation and EG-0 (6)
The foot to foot current path should not be
automatically ruled as an ‘insignificant risk’
for this current path is the one most
commonly encountered in practice and
invariably experiences the maximum
exposure time, meaning that under some
circumstances, the level of risk posed by
this mechanism could prove statistically
significant.
we_n7071236_v1.ppt 24
AGGREGATE RISK TO BUSINESS (1)
we_n7071236_v1.ppt 26
AGGREGATE RISK TO BUSINESS (3)
Embedded processes and procedures must also
clearly demonstrate that all of the initial and
ongoing safety coordination needs that involve
other licensed operators, and large industrial
enterprises that are dependent on the reliable
supply of electric power, are adequately applied.
This requirement is not trivial and will be
particularly challenging when large industrial
entities are involved, such as the Kwinana
Industrial corridor in WA, for the holistic
assessment and apportionment of the safety
risks posed is expected to be complex.
we_n7071236_v1.ppt 27
AGGREGATE RISK TO BUSINESS (4)
we_n7071236_v1.ppt 28
RISK ASSESSMENT DATA AND
ROLE FOR REGULAR AUDITING (1)
The recorded material needs to be logically
structured, regularly updated and made as
transparent as possible in order to meet the
need of regular internal and periodic external
auditing.
The material should be securely controlled and
managed in order to respond to other (e.g. legal
or coronial) enquiry purposes. Most important is
that there is no ambiguity within the business as
to what the accountabilities are for the upkeep
and maintenance of these aspects.
we_n7071236_v1.ppt 29
RISK ASSESSMENT DATA AND
ROLE FOR REGULAR AUDITING (2)
It is recommended that annual internal audits be
undertaken with summary reporting made a
standing order to those sectors of management
holding the accountability for duty of care.
It is anticipated that relevant regulating bodies
will want to conduct periodic independent audits,
at the end of an access agreement period for
example, to verify the safety performance etc of
the licensed utility is to an acceptable standard,
as well as to ensure that the Regulator(s) own
due diligence responsibilities are being
adequately met.
we_n7071236_v1.ppt 30
PART 2
by
Dr Zoran Božić
we_n7071236_v1.ppt 31
PART 2
Critical review of individual building blocks:
• Probability of coincidence calculation
• Probability of fatality calculation
• Assessment of risk due to step voltages
• Electric shock duration - autoreclosing &
credible contingencies
• Table of permissible risk limits
we_n7071236_v1.ppt 32
PROBABILITY OF COINCIDENDE
CALCULATION (POCC) (1)
Unexpected outcomes of draft EG-0
Origin of unexpected outcomes
Link between Eq.(A-13) and the
“Probability That Two Trains Meet at the
Station” problem
“Two Trains…” inappropriate model
What should be done instead
The magnitude of error
we_n7071236_v1.ppt 33
POCC – Equation (A-13) (2)
we_n7071236_v1.ppt 35
POCC – Unexpected Outcomes (4)
we_n7071236_v1.ppt 36
POCC – Origin of Unexpected
Outcomes (5)
Partly originate from approximations made
in deriving Eq.(A-13) in section A.1.1
The main cause is use of the inappropriate
mathematical model of the “Probability
That Two Trains Meet at the Station”
problem for the intended electrocution
application.
Moreover, two different solutions to that
problem are given in EG-0
we_n7071236_v1.ppt 37
POCC – Origin of Unexpected
Outcomes (6)
One is the approximate analytical solution
of section A.1.1, ending with Eq.(A-13)
The second is the graphical derivation of
section A.1.2(A), ending with Eq.(A-16)
Their comparison reveals that Eq.(A-13) is
essentially the first term of Eq.(A-16)
The 2nd term, truncated, ensures that the
probability does not exceed the value of 1
we_n7071236_v1.ppt 38
POCC – Link Between Eq.(A-13)
and “Two Trains …” (7)
Definition of events A and B in section
A.1.1:
“Event A – a person in contact with an
earthed asset when a fault occurs”,
“Event B – a person contacts an earthed
asset during a fault”
The ‘Two Trains” are the ‘fault coming’
train and the ‘person coming’ train
we_n7071236_v1.ppt 39
POCC – Link Between Eq.(A-13)
and “Two Trains …” (8)
we_n7071236_v1.ppt 40
POCC – “Two Trains …”
Inappropriate Model (9)
The value calculated by Eqs.(A-13 & 16) is the
‘cumulative’ probability – the summation of
probabilities of a number of events
Note that all but one of these events represent
electrocutions of shorter durations than the
protection clearance time
Consequently, the solution to the “Two Trains…”
problem does not represent any single real
event of specific duration
we_n7071236_v1.ppt 41
POCC – “Two Trains …”
Inappropriate Model (10)
we_n7071236_v1.ppt 42
POCC – What Should be done
Instead (11)
The electrocution event of duration equal
to the protection clearance time should be
used for the earth grid design
The probability of a single event i is given
in Eq.(A-14) as:
Pcoincidence_i = (fd_i x pd_i) / T2
and also illustrated in Figure A-1(c) of
Draft EG-0, Feb 2010
we_n7071236_v1.ppt 43
POCC – The Magnitude of the
Error (12)
we_n7071236_v1.ppt 45
POCC – The Magnitude of the
Error (14)
we_n7071236_v1.ppt 46
PROBABILITY OF FATALITY
CALCULATION (1)
Pfatality = Pcoincidence x Pfibrillation
Rewritten for a single event i becomes:
Pfatality_i = Pcoincidence_i x Pfibrillation_i
we_n7071236_v1.ppt 49
ASSESSMENT OF RISK DUE TO
STEP VOLTAGES
we_n7071236_v1.ppt 50
ELECTRIC SHOCK DURATION (1)
we_n7071236_v1.ppt 52
ELECTRIC SHOCK DURATION –
USA (3)
we_n7071236_v1.ppt 53
ELECTRIC SHOCK DURATION –
CREDIBLE CONTINGENCIES (4)
Some argue that safety and other
assessments should be done consistently
as per the list of credible contingencies
Earth faults cleared in the backup
protection clearance time can be a
credible contingency
This also implies that the same time
should be used for safety and equipment
rating, and other power system studies
we_n7071236_v1.ppt 54
ELECTRIC SHOCK DURATION -
AUTORECLOSING (5)
Autoreclosing, particularly high-speed,
result in longer clearing times increasing
the level of risk to public and personnel
Can this additional risk be ignored?
……
The proposed thresholds in EG-0 originate
from IEC 60479 – I/t curves
They relate to ventricular fibrillation
we_n7071236_v1.ppt 55
ELECTRIC SHOCK DURATION -
AUTORECLOSING (6)
we_n7071236_v1.ppt 56
ELECTRIC SHOCK DURATION -
AUTORECLOSING (7)
“It is important that the thresholds as order
of magnitude are valid for all persons
(men, women and children) independent
of their state of health. Often concerns are
expressed in that respect but if the
background of such objections is
examined it is found that such objections
represent just opinions without
experimental evidence….”
we_n7071236_v1.ppt 57
ELECTRIC SHOCK DURATION -
AUTORECLOSING (8)
“For fault durations below 200 ms,
ventricular fibrillation is only initiated within
the vulnerable period if the relevant
thresholds are surpassed.….”
Autoreclosing causes, as a minimum, two
successive electric shocks of the
aggregate duration of twice the primary
protection time, which is generally longer
than 200ms
we_n7071236_v1.ppt 58
ELECTRIC SHOCK DURATION -
AUTORECLOSING (9)
we_n7071236_v1.ppt 59
ELECTRIC SHOCK DURATION -
AUTORECLOSING (10)
It is important to note that the total duration
of electrocution is (implicitly) considered in
Figure 20, regardless if it is ‘administrated’
as a ‘single large dose’ (one electrocution
of longer duration, say due to a circuit
breaker fail time) or in ‘two consecutive
smaller doses’ (the sum of two primary
protection clearing times, which is typical
for high speed autoreclosing, with only one
autoreclosing attempt).
we_n7071236_v1.ppt 60
ELECTRIC SHOCK DURATION -
AUTORECLOSING (11)
we_n7071236_v1.ppt 61
ELECTRIC SHOCK DURATION –
USA (12)
IEEE Std. 80 states that the cumulative
effect of two or more closely spaced
shocks is assessed by using the sum of
individual shock durations as the time of a
single exposure:
“The fault duration and shock duration are
normally equal, unless the fault duration is
the sum of successive shocks, such as
from reclosures.”
we_n7071236_v1.ppt 62
ELECTRIC SHOCK DURATION -
UK (13)
In UK, a time about twice the primary
protection clearance time is used
200ms (as a minimum), where the fastest
clearing times are 80ms at 400kV, 100ms
at 275kV and 120ms at 132kV; the longest
operating time to clear a transmission fault
is 140ms (including 3 ended lines)
500ms in the distribution system
we_n7071236_v1.ppt 63
ELECTRIC SHOCK DURATION -
Europe (14)
In a number of other European countries
in which high speed autoreclosing is
routinely used, as it is in WA, it is also
required that the two (primary) protection
clearing times (before and during
autoreclosing) be added and their sum
used for safety assessment.
For example, 2x150 = 300 ms
we_n7071236_v1.ppt 64
ELECTRIC SHOCK DURATION –
AS2067 (15)
The earlier version of AS2067, which was
referencing to IEEE Standard 80, also
required to add up the (two) protection
clearance times and use the sum for
safety assessment (where autoreclosing
applies).
This fact and the associated change may
have been overlooked in the development
of the AS2067-2008 version?
we_n7071236_v1.ppt 65
TABLE OF PERMISSIBLE RISK
LIMITS (1)
An important characteristic of the
approach proposed in Draft EG-0 is that
the calculated probability of fatality is
compared against the thresholds which
are expressed as absolute values.
Central for success of this new approach
is that the probability is calculated
accurately and that the appropriate
thresholds are selected.
we_n7071236_v1.ppt 66
TABLE OF PERMISSIBLE RISK
LIMITS (2)
This section examines the thresholds and
challenges their suitability` in the present
form by raising the following issues:
• Absolute values do not apply universally
• Impact of the neighboring land use on the
“Low or Tolerable Risk” threshold
• The impact of apportion of the aggregate risk
we_n7071236_v1.ppt 67
TABLE OF PERMISSIBLE RISK
LIMITS (TOPRL) (3)
we_n7071236_v1.ppt 68
TOPRL – Absolute values do not
apply universally (4)
• Largely originate from the Environmental
Protection or similar agency documents
• The maximum aggregate risk permissible to
be exported to outside of the installation
• The values vary for different land uses
• Apportion to individual industries = ?
• Allocation of a fraction of that aggregate risk
would require more stringent earth grid design
we_n7071236_v1.ppt 69
CONCLUSION (1)
Some short comings have been explained
in the approach used to assess and
quantify the risk exposure probability and
an alternative approach was presented
It is recommended that the calculation of
Pfatality be fully completed for all situations
that require an in-depth analysis and not
be truncated on the basis that Pcoincidence is
shown to be “≤ 10-6”
we_n7071236_v1.ppt 70
CONCLUSION (2)
All relevant system performance statistics
and fault clearing times, etc should be
included as a matter of routine as part of
the risk management data base
we_n7071236_v1.ppt 71
CONCLUSION (3)
Under Western Australian (Supply and System
Safety) Regulations 2001 there is the potential
for serious conflict between AS2067–2008 and
EG-0 over the interpretation and use of V/t
curves and in this context AS2067–2008 is seen
to be ambiguous.
Greater clarity and consistency would be
achieved if AS2067–2008 were to reference the
current/time zone curves presented in IEC
60479-1 rather than the V/t curves currently
shown in this standard.
we_n7071236_v1.ppt 72
CONCLUSION (4)
Little consideration is presented on the matter of
how to deal with the aggregate risk associated
with large industrial complexes and the complex
interrelationships that can affect the level of
societal risk.
While this is seen to be primarily a responsibility
for WA, and possibly National, regulators to
clarify and coordinate, some comments to this
effect in EG-0, assuming this to be so, would
beneficial.
we_n7071236_v1.ppt 73
Questions = ?
we_n7071236_v1.ppt 74