INTRODUCTION
Arisk management approach can provide a decision-making
method for preventive conservation of
a museum collection or group of collections. Risk
is the chance of an undesirable change occurring.
Risk assessment is the analysis of the magnitude of
each and all risks affecting some entity. Risk management
is the application of available resources
in a way that minimizes overall risk (Crouch and
Wilson, 1982; Moore, 1983; Suokas and Rouhiainen,
1993).
We make dozens, if not hundreds, of risk assessments
and risk management decisions each day.
Examples of risk management decisions that many
of us make daily before we even start work include
deciding whether an extra five minutes of sleep will
result in being late for work, determining when a
hot cup of coffee has cooled enough to be sipped
safely, and judging whether the time saved by
exceeding the speed limit while driving to work is
worth the risk of an accident or fine.
Assuming that we get up and arrive at work
safely, we continue to make risk assessment and
management decisions as we work with collections.
Many decisions are exceedingly simple. An
example is deciding whether to re-close the cap on
a jar of fluid preservative after a specimen has been
removed for a brief examination, or to leave it open.
A somewhat more complex decision is judging how
many small specimens can be carried safely at one
time. Occasionally, major decisions must be made.
An example is deciding if a humidification system
should be installed in a temporary storage facility
that has existed for several years without humidification.
Finally, at least once each year, anyone
involved with budget allocation must decide how
limited resources can be distributed to best reduce
risk to collections.
Some of these decisions are so simple and
repetitive that they can become a procedure.
For instance, a standard procedure could be to
re-close a container of fluid preservative before
one minute has elapsed. Other decisions are less
repetitive and will always require some judgment.
For example, twenty small specimens
might be carried safely on a tray if each specimen
weighs only a few grams, while it could be dangerous
to carry more than a single specimen if each
specimen weighs a few kilograms. Other decisions,
such as the installation of a humidification system
in a temporary storage area, are complex. A wise
decision requires consideration of both immediate
and long-term risks associated with the status quo,
the proposed action, and alternative methods of
mitigating the problem, such as humidity control
by enclosure and buffering.
Finally, determining how to use limited resources
to obtain the greatest possible reduction in
total risk to collections is a very complex decision.
It requires the assessment of so many types of risk
and possible means of mitigation that it cannot be
made properly without an organized method of
evaluating the relative costs and benefits of alternative
strategies.
A risk management approach can be used, not
only to organize thoughts on any decision affecting
the preventive conservation of collections, but also
to provide a method for considering the most difficult
decision we face - how limited resources can
best be applied to the protection of collections.
A RISK MANAGEMENT APPROACH
A risk management approach to collection preservation
issues involves four basic steps:
1) identifying all risks to collections,
2) assessing the magnitude of each risk,
3) identifying possible mitigation strategies, and
4) evaluating the costs and benefits associated
with each strategy.
These steps are discussed in a general way to
make them applicable to all types of collections.
IDENTIFYINGALLRISKS TO COLLECTIONS
The application of risk management to preventive
conservation has been made possible by the
development of a comprehensive classification
of the agents of deterioration (Michalski. 1987,
1990b). Nine agents of physical deterioration,
and one agent of non-physical deterioration (custodial
neglect), are listed in column 1 of Table 1.
RISK MANAGEMENT APPLIED TO PREVENTIVE CONSERVATION
Robert R. Waller
22
STORAGE OF NATURAL HISTORY COLLECTIONS: A PREVENTIVE CONSERVATION APPROACH
Table 1. Specific examples of types of risk and the relative importance of implementing means of control
at each possible level for control. Type of risk: 1 = catastrophic, 2 = severe, and 3 = mild/gradual
(see text explanation for type of risk).
23
WALLER - RISK MANAGEMENT APPLIED TO PREVENTIVE CONSERVATION
Figure I. The ranges of frequency and severity of
the types of risk(1, 2, and 3).
Each agent of deterioration can be manifested
as one or more of three different types of risk,
characterized by the frequency of occurrence and
the severity of effect on collections. This is shown
schematically in Figure 1. Risks that belong in the
lower right section of this diagram, those tending to
be rare in occurrence and gradual in effect, are so
insignificant that they are of little concern. The risk
of damage due to exposure to incorrect temperature
in a collection of diamonds is an example. Risks
that belong in the upper left section of the diagram,
those tending to be constant in occurrence and
catastrophic in effect, are so severe that collections
subject to them are unlikely to exist. A snowball
collection in a hot room is an example. Three types
of risk remain to be considered:
1) rare and catastrophic,
2) sporadic and severe, and
3) constant and mild/gradual.
The indicated limits in frequency and severity for
each type of risk are arbitrary and may vary among
individual collections. An example of a Type 1 risk
is fire, which occurs only rarely but has a catastrophic
effect. At the other end of the spectrum,
an example of Type 3 risk is damage caused by low
levels of pollutants in collection areas. It occurs at
a nearly constant rate and is mild and gradual in
effect. Examples of different types of risks associated
with agents of deterioration are given in
column 3 of Table 1.
ASSESSINGTHEMAGNITUDEOFEACHRISK
The purpose of a risk assessment is to make commensurable
the risks to collections that accrue
from various threats, ranging from earthquakes
at one extreme to slightly warmer than optimal
temperatures at the other. The basis for this commensurability
is the combination of probability and
severity factors. All risks, regardless of type, have
a magnitude that is the product of probability and
severity (see Table 2). In practice, for most risks
that are not Type 1 (rare and catastrophic in nature),
extent is considered instead of probability. Severity
is interpreted to be the product of the fraction of the
collection susceptible to the threat and the potential
loss in value to the collection (Waller, 1994).
Frequently, a range of probabilities can be
used to describe the severity of a possible occurrence;
making precise calculation of the risk more
complex. For example, a range of probabilities
describes the chance of a fire that could result in
complete, partial, or minimal damage to a collection.
While progress is being made in dealing with
these complexities (see, for example, Harmathy et
al; 1989) approximations that can be obtained by
considering extremely simplified models are useful
as starting estimates (Waller, 1994).
Similarly, the total risk to a collection from all
agents of deterioration can be calculated as the
combination of all individual risks. However, the
total risk is not the simple sum of the individual
risks. A complete description of the combinative
analysis required to determine a numeric value for
the total risk to a collection is beyond the scope of
this chapter. Further, at present, this exact calculation
is probably of little value considering the
uncertainties in our knowledge of the magnitudes
of individual risks.
Nevertheless, a semiquantitative understanding
and application of this risk management procedure
will lead to better decisions on preventive conservation
issues. In addition, by attempting to estimate
the magnitude of individual risks, we can identify
the risks for which reasonable estimates of probability
or severity cannot be made. This procedure,
therefore, also identifies the factors affecting collection
preservation that are most in need of study.
Taking all of these factors into consideration,
it is still possible and advantageous to attempt to
estimate the magnitude and the bounds of uncertainty
of the estimate for each identified risk to
a collection. The classification of risks into three
types is helpful in establishing these estimates
because sources of information on the probability
and severity associated with each risk depend on
its type.
As a rule, information on Type 1 risks (infrequent
disasters) must be obtained from central
24
STORAGE OF NATURAL HISTORY COLLECTIONS: A PREVENTIVE CONSERVATION APPROACH
agencies such as geological surveys, flood prediction
authoritie, and insurance data bureaus. The
agency best equipped to provide this information
will depend on the political jurisdiction and geographic
area in which the collection resides. Museums
administered under a parent organization may
have access to a risk management expert employed
by the organization. If so, the risk management
expert could assist in interpreting information on
Type 1 risks.
Information on Type 2 risks (intermediate frequency),
such as damage from mishandling, must
come from conservation documentation. The magnitude
of Type 2 risks can be determined precisely
only through analysis of condition data over time.
This fact further supports the need for collection
and specimen condition reports and suggests some
of the information that these reports must record.
The magnitude of Type 3 risks (constant) should
come from conservation science studies coupled
with monitoring of environmental conditions in
collections. A model of our ability to predict these
rates is exemplified in the photo degradation studies
of many materials (Michalski, 1990a). Further studies
of the rates of deterioration of materials exposed
to specific ranges of environmental conditions are
needed to increase the number of deterioration
processes for which we can predict rates. Until sufficient
information is available from such studies,
conservation documentation can provide some idea
of the magnitude of these risks for a particular col-
lection.
IDENTIFYINGPOSSIBLE MITIGATION STRATEGIES
Means of Control
After all risks to a collection have been identified
and their magnitude quantified, or at least estimated.
risk mitigation strategies can be explored.
There are three basic means for mitigating a risk:
1) eliminate the source of the risk,
2) establish a barrier, and
3) act on the agent responsible for the risk,
Often all three means can be used to reduce a risk.
but one may prove to be most effective.
Eliminate source
The risk of collection damage from an incorrect
(too high) relative humidity level in a basement
storage area can serve as an example of how each
strategy could be applied to the same problem. If
the problem is caused by infiltrating damp, resulting
from poor drainage around the building, then
eliminating the risk might involve installing gutters
along eaves and installing drainpipes running
to dry wells at a distance; grading the ground level
away from the building; installing weeping tile
drains around the foundation; or some combination
of these. Alternatively, if the problem is a result of
the basement remaining cooler than humid outside
air, then heating the basement area could eliminate
the risk from incorrect relative humidity. In most
cases, eliminating the threat is the preferred means
of control. However, in the example above, heating
the basement area may create or increase other
threats to the collections, such as thermal degradation,
or fire, and hence may be inadvisable.
Establish barrier
Establishing a barrier is usually the next preferred
means of control. In the example above,
where infiltrating damp was the source of the problem,
applying waterproof coatings to the foundation
may adequately reduce the relative humidity
and eliminate the risk. If the problem results from
low temperatures in the basement, then the problem
may be a seasonal one. Assuming that the annual
average relative humidity is appropriate for the collection,
the best barrier might be created by using
well- sealed, water-vapor-impermeable cabinets
for the collection. If this option were adopted, the
procedures for working with the collection might
have to be modified to ensure that specimens would
nor be removed from cabinets for an extended time
during periods of high relative humidity.
Act on agent
Acting on the agent responsible for the risk is
usually done when the other means of control have
failed to reduce the risk sufficiently. For the above
25
WALLER - RISK MANAGEMENT APPLIED TO PREVENTIVE CONSERVATION
example, installation of a dehumidifier would be
considered acting on the agent. It is worth noting
that while this direct approach is the one that often
would be considered first. depending on the source
and extent of the problem, it could prove to be the
worst choice when all long-term costs and risks are
considered. In this case, the presence of a dehumidifier
could increase the risks of fire and of local
flooding, particularly if maintenance and servicing
requirements could not be met.
All possible means of control should be considered
for mitigating each significant risk to collections.
Frequently, one of the methods will be
most appropriate and provide the best cost/benefit.
In other cases, effective control of the risk might
require the use of all three methods together.
Levels for Control
Each of the three basic means of control
described above can and should be considered at
each of the eight possible levels for control: location,
site, building, room, cabinet, specimen, policy,
and procedure.
Location
The location of collections often cannot be controlled.
If, however, location is a matter of choice,
then all possible care should be taken to ensure that
it minimizes exposure to both natural and humanmade
risks.
Site
Many risks to collections can be significantly
affected by modification to the building site. As
noted above, the risk from an incorrectly high relative
humidity might be reduced by site modifications
to improve drainage.
Building
Buildings and building-wide systems traditionally
constitute the level on which museums rely
most heavily for control of risks to collections. This
is certainly an important level for controlling risks
from most agents of deterioration. However, estimation
of risks for which building-wide systems
are a major means of mitigation must consider that
such systems will probably fail to operate or may
malfunction over a projected period.
Room
Means of control applied at the level of individual
rooms are also important for control of risks
associated with most of the agents of deterioration.
One such example is to install light switches for
each collection storage room or, better still, for
discrete areas within a room in order to reduce
light-induced deterioration in collections.
Cabinet
The importance of cabinets in reducing risks to
collections cannot be overemphasized. Cabinets
can provide effective barriers to many agents of
deterioration, especially pests, light, and relative
humidity fluctuations. Although initial cost for
high-quality cabinets is significant. maintenance
costs are very low and useful life is long. Therefore,
cabinets become very cost effective over time.
Specimen
The specimen level of risk mitigation is of
primary importance to the preservation of fluidpreserved
specimens. The quality of the seal on
each specimen container is critical for reducing
oxidation, which increases acidity, and evaporation.
which results in dilution of alcohol strength. For
other types of collections, specimen-level means
of control include adequate physical support, dust
covers, and additional protective measures.
Policy
The policy level of risk mitigation is especially
important for reducing risks from custodial neglect.
For example, needless damage to specimens from
inappropriate use can be eliminated by establishing
and enforcing a policy that defines appropriate
uses.
Procedure
Finally, proper and well-established procedures
are essential to an effective overall risk management
strategy. In many cases such procedures will,
by themselves, provide the most cost-effective
manner of reducing a risk.
EVALUATINGCOSTSANDBENEFITSASSOCIATEDWITH
EACHSTRATEGY
After all risks to collections have been identified
and quantified, at least by rank, and all possible
means of control applicable at any level have been
identified, an assessment is made of the cost and
benefits associated with each strategy. This assessment
should be guided by the considerations discussed
below.
Effect of a Proposed Strategy on Each and
All Agents of Deterioration
Commonly, a strategy to reduce the risk from
one agent of deterioration will influence the risk
from another agent. For example, spray humidifiers
26
STORAGE OF NATURAL HISTORY COLLECTIONS: A PREVENTIVE CONSERVATION APPROACH
installed without adequate water treatment may
pose more problems than the low humidity levels
that they solve because of particulate pollutant
accumulation on collections (Rogers and Costain,
1980). More optimistically, strategies such as the
use of quality cabinets may reduce many risks
from a range of agents of deterioration, including
criminals, light, contaminants, and incorrect relative
humidity.
Cost/Benefit Associated with Both
Implementation and Maintenance Stages
Most people have a tendency to carefully consider
the costs associated with implementation or
installation but not those associated with maintenance.
Numerous examples can be cited where
climate control systems failed to function according
to specifications because there was inadequate
support for required maintenance.
Conversely, most people have a tendency to
think about the future benefits of a risk reduction
strategy and neglect to consider, or underestimate,
the risks to which collections may be
exposed while the strategy is being implemented.
This is especially important when the strategy
is expected to reduce risks for a relatively limited
period. An example is modifications to temporary
facilities. In this case, increased risks involving
physical forces, fire, criminal acts, and pollutants
(for example, dust) during renovation may exceed
the anticipated reduction in risk over the time the
collection is in the temporary facility.
Effect of a Proposed Strategy on Risks
Other Than Those to Collections
Throughout this chapter, the focus has been
on risks to collections and the evaluation of the
effect of mitigation strategies on those risks alone.
However, before a strategy is implemented, it is
essential to consider the effect of the strategy on
risks other than those to collections. One must
consider, for example, the effect of the strategy on
the health and safety of staff, visitors, and users of
the collection. As a more complex instance, the
risk to the local and global environment introduced
by installation of an energy-intensive airhandling
system should be considered.
SUMMARY
A method for evaluating risk to collections and
determining which risk mitigation strategies would
provide the best cost/benefit ratio in reducing total
risk to collections has been outlined. At present, the
precision with which this method can be applied
is limited by a lack of derailed information on the
magnitude of many of the risks affecting collections.
It is evident, however, that order of magnitude
estimates for individual risks can provide
improved understanding of the size and nature of
risks to collections. Further, the sources that need to
be developed and exploited to provide more accurate
information have been identified for each type
of risk. This information helps to establish goals in
both conservation documentation and conservation
science.
The risk management approach to conservation
planning and decision making offers several other
advantages. The approach is holistic in nature. All
sources of risk are considered, and the scope of the
assessment can be made broad enough to include
the impact of proposed actions on the global environment.
The solutions developed through this
method are necessarily pragmatic because they
will provide the greatest reduction in risk to collections
for any given amount of resources available
for preventive conservation. Finally, by providing
a clear basis for requesting resources to mitigate
risks, this approach will improve the credibility of
the request and, consequently, increase the probability
of acquiring the resources.
ACKNOWLEDGMENTS
I am grateful to Stefan Michalski, Canadian
Conservation Institute, Department of Communications,
for generously sharing his experience,
ideas, and insight, and to Sylvie Marcil, Canadian
Museum of Nature, for suggestions for improving
the chapter.
LITERATURE CITED
Crouch, E. A. C. and R. Wilson. 1982. Risk/Benefit
Analysis. Ballinger Publishing Company, Cambridge
MA, xiii + 218 pp.
Harmathy, T. Z. and Delphi Group. 1989. A decision
logic for trading between fire safety measures.
Fire and Materials, 14:1-10.
Michalski. S. 1987. Preventive conservation draft
poster, Draft 1. May/87.
_______.1990 a. Towards specific lighting guidelines.
Pp. 583-588 in Preprints of the International
Council of Museums, Committee for Conservation,
9th Triennial Meeting, Dresden, 26-31 August
1990. ICOM Committee for Conservation, Los
Angeles, 840 pp.
27
WALLER - RISK MANAGEMENT APPLIED TO PREVENTIVE CONSERVATION
_______.1990 b. An overall framework for preventive
conservation and remedial conservation. Pp.
589-591 in Preprints of the International Council
of Museums, Committee for Conservation, 9th
Triennial Meeting, Dresden, 26-31 August 1990.
ICOM Committee for Conservation, Los Angeles,
840 pp.
Moore, P. G. 1983. The Business of Risk. Cambridge
University Press, Cambridge. x + 244 pp.
Rogers, G. DeW. and C. G. Costain. 1980. Contamination
of a collection by a demineralizer/aerosol
humidifier system. Journal of the International
Institute for Conservation-Canadian Group, 5(1/2):
22-24.
Suokas, J. and V Rouhiainen, 1993. Quality Management
of Safety and Risk Analysis. Elsevier,
Amsterdam, 291 pp.
Waller, R. 1994. Conservation risk assessment:
A strategy for managing resources for preventive
conservation. Pp. 12-16 in International Institute
for Conservation, Preprints of the Ottawa Congress,
September 1994. International Institute for
Conservation, London. 244 pp.
AUTHOR
Robert Waller is Managing Director, Collections
Division, Canadian Museum of Nature. Mr.
Waller has published numerous articles on the conservation
of mineral specimens and fluid-preserved
collections. He is also past President of the Canadian
Association of Professional Conservators.
Robert R. Waller
Collections Division
Canadian Museum of Nature
Ottawa. Ontario KIP 6P4
Canada