Geotechnical News • September 2018
37
GEOHAZARDS
approving officer for the Province, Mr.
Elston, refused to allow plan deposi-
tion, and thus no title was conveyed
to the subdivision, because he argued
that the development of the subdivi-
sion would be against the public inter-
est. Mr. Elston’s decision rested on the
potential of a catastrophic landslide
originating at the Barrier to reach the
development. An appeal was launched
by Cleveland Holdings against this
decision. Judge Berger presided the
proceedings and made some key
conclusions:
“Dr. Mathews and Mr. Naismith both
calculate the risk of a [catastrophic]
slide on a time scale of thousands of
years. They say there is a probability
of a slide at the Barrier in the next
10,000 years. It may occur next year,
it may occur in a thousand years, it
may occur in 10,000 years. Yet for
both of them the risk is real enough
that neither would want to live in
the subdivision. The risk is one they
would prefer to avoid.”
In his decision, Judge Berger followed
the logic of a risk analysis. He care-
fully examined the meaning of prob-
ability and identified that the hazard
and risk is quantifiable and real. His
final judgement, therefore, may consti-
tute the first risk-based court decision
with regard to landslide hazards in
British Columbia.
This case, later supported by the
Garibaldi Advisory Panel (1978),
demonstrated the reluctance of the
judge to accept a risk, which at the
time was poorly quantified. Given that
a catastrophic failure had occurred in
the past 120 years, he was not willing
to allow development to proceed.
This decision is key in the develop-
ment of the 1:10,000-year probability
or return period “standard” that has
been propagated by Cave’s (1993)
work and has since been manifested
in the Guidelines for Subdivision
Approval Officers (MoTI 2015) as
well as criteria that are being used to
this day by the Fraser Valley Regional
District in managing their geohazards.
Notably, however, Judge Berger’s
decision does not state explicitly that
the 1:10,000-year event should be the
target landslide probability or return
period at which subdivisions are not
to be approved. Rather, by analogy to
the post glacial, Holocene Epoch, he
was referencing a 10,000-year sample
frame.
EGBC (2012) recognized the difficulty
in reliably estimating the 1:10,000-
year event for debris flows, and
oriented itself on the existing national
seismic code, which stipulates the
use of an approximately 1:2,500-year
event in building design. The underly-
ing logic is that buildings designed to
withstand ground motion accelerations
commensurate with a 1:2,500-year
earthquake would allow safe building
exit, while a 1:10,000-year event may
not. Implicit is the recognition that
building design for a 1:10,000-year
event may be cost prohibitive and that
some level of unquantified residual
risk is therefore tolerable. For land-
slide hazard and risk assessment and
mitigation design, the 1:2,500-year
event is still a threshold beyond those
typically used in international practice
for landslides.
In comparison, the EGBC Guidelines
for Legislated Flood Assessments in
a Changing Climate in BC (2012)
classify assessments by the number of
buildings proposed within a subdi-
vision and suggest variable return
periods up to the 1:2,500-year event
be considered for flood hazard assess-
ments and mitigation design.
In contrast, the Canadian Dam Asso-
ciation (CDA) and similar associa-
tions around the world classify dams
according to their potential conse-
quences of failure, and recommended
that assessment and design criteria
vary accordingly. For example, CDA
classifies dams and assigns annual
exceedance probabilities for earth-
quakes and floods for consideration
and consultation in a risk-informed
design and management considering
up to the 1:10,000-year event where
dam failure could result in “notionally
more than 100 expected fatalities”.
Proposed changes
This article highlights some incon-
sistencies in how natural hazards are
being addressed through existing guid-
ance and regulations. The following
points attempt to provide improve-
ments to the present situation, though
the authors cannot claim to provide
lasting solutions that may require site-
specific stakeholder engagement.
• The approach of natural hazard risk
management requires homogeniza-
tion in BC. This does not imply
application of the same range of
frequencies for each geohazard,
but a recognition of scientifically
defensible reconstructions and
statistically valid extrapolations or
projections. This includes recogni-
tion of the highly non-stationary
paraglacial decline in geomorphic
activity during the Holocene.
• Competing guidelines should be
either amended or abandoned. The
key is to create a single broad con-
sensus document applicable by all
ministries, local governments and
industries which ought to undergo
detailed review by practitioners,
academics and regulators alike,
and which would be subject to
episodic review, refinements and
updates.
• The range of hazard frequency
should be hinged to the level of
proposed development density.
This is in line with the notion of
group- or societal risk. Very rare
events can and will lead to fatali-
ties. Society is willing to assume
some risk for the perceived ben-
efits of living in a desirable area.
All future updates and refinements
of EGBC landslide guidelines
should include guidance on land-
slide safety criteria for landslide
types other than only debris flows
and debris floods.
• Specifically, for locations with a
credible catastrophic landslide
hazard the use of Quantitative Risk
