PREVENTIVE CONSERVATION PLANNING

FOR LARGE AND DIVERSE COLLECTIONS

ROBERT WALLER

ABSTRACT -The Canadian Museum of Nature is in the process of implementing three

systems to enable it to plan the use of, and be accountable for, resources directed toward
preventive conservation. The first system, risk assessment and management, employs
simplifying models to obtain estimates of the magnitudes of specific risks to collections.
Based on these, mitigation strategies are proposed and evaluated in terms of costs, risks,
and benefits both during the implementation phase and over a projected period of time.
The system of categories of specimens facilitates, when necessary, setting priorities
for risk mitigation projects on the most important parts of collections. Finally, specific
preventive conservation issues that are not best quantified from the broad perspective
of risk assessment and management are addressed by a collection profiling system.
This system was derived from the collection health index system introduced by R.
McGinley. It identifies preventive conservation issues arising within demarcated parts
of collections.

1. INTRODUCTION
Over the past two decades considerable progress has been made in making the
planning of preventive conservation more systematic and comprehensive (Wolf, 1993;
Michalski 1990a, 1994). At present, conservation plans developed in accordance with
recognized methodologies will be reasonably comprehensive. Alas, being comprehen-
sive is no guarantee of being effective and certainly no guarantee of being cost-effective.
This is true because most of these methodologies are based on inventories or checklists
of control systems in place.
Several years ago the Canadian Museum of Nature (CMN) became completely
pragmatic in planning for preventive conservation. We are doing this by using risk
assessment and management methods coupled with a recognition of a hierarchy of
value among objects in collections. This was done knowing well that all of the informa-
tion required to do this comprehensively and accurately was not currently available.
Our several years of experience have proven to us that, despite the lack of reliable
information for precisely quantifying all risks and assigning values to all objects, our
investment in preventive conservation is now much more effectively focused than it
would otherwise be.

2. RISK ASSESSMENT AND MANAGEMENT

Our risk assessment and management systems have been described elsewhere
(Waller 1994, 1996) and training in their application is available (Canadian Museum of
Nature, 1994). The following is a review of the main concepts.

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 1

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints
2.1 RISK ASSESSMENT
Risk of loss and damage to collections arises from exposure to one of ten agents
of deterioration. These are: physical forces, fire, flood, criminals, pests, contaminants,
light and UV radiation, incorrect temperature, incorrect relative humidity, and custodial
neglect. Most of these are self-explanatory. The last, custodial neglect, includes a variety
of intellectual or legal shortcomings that lead to loss of value of objects, of objects
themselves, or of their data through a variety of means including physical misplacement,
failure to secure legal title, failure of information systems, and so on.

Risks vary both in frequency of occurrence and severity and it is important that
this be recognized. Although this variation is continuous, we have found it useful to
define three types of risk. These are: 1) rare and catastrophic, 2) sporadic and intermedi-
ate in severity, and 3) constant and mild/gradual. Many agents of deterioration present
risks of all three types. For example, the agent physical forces includes earthquakes,
dropping a drawer or crate of objects, and distortion due to improper support. Others,
such as fire, type 1, and light damage, type 3, occur only as a single type of risk.
Recognizing the distinctive nature of these three types of risk clarifies the idea that
different kinds and sources of information are required for estimating magnitudes of
each type of risk.

The combination of ten agents of deterioration and three types of risks leads
to 23 useful categories of risk that are useful to consider. These 23 categories of risk
are not only comprehensive but also are easily seen to be comprehensive. This is a
great benefit when approaching senior management, governing bodies and grant-giving
organizations with a plan for reducing risks. They are all more inclined to support a plan
that addresses an issue in such a comprehensive manner. Although it is possible that
priorities may shift as new information becomes available, there will be no completely
new issues raised.

The magnitudes of risks over a one hundred-year forecast period are estimated
through application of simplifying models to evaluate four parameters. These are: Frac-
tion Susceptible (FS) interpreted as the part of the collection that is potentially
subject to loss or damage by exposure to the risk being considered. Loss in Value (LV)
defined as the maximum reduction in value, usually in a utilitarian rather than a mon-
etary sense, resulting from exposure of the fraction susceptible to the risk being con-
sidered. Probability (P) defined as one for type 2 and type 3 risks, which are certain
to occur, and evaluated, in conjunction with extent, for type 1 risks. This evaluation
requires expertise from professionals in fields such as seismic engineering, fire protec-
tion, and so on. Extent (E) is the measure to which a risk is expected to produce the
defined Loss in Value to the Fraction Susceptible over the forecast period.

The Magnitude of Risk, for each specific risk identified, is then calculated as
the product: MR = FS x LV x P x E. This expression differs slightly from that given in
earlier references (Waller, 1994, 1996). P and E are now resolved parameters rather than
a blended parameter.

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 2

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints
 Information required to estimate magnitudes of risks is becoming increasingly
available for all types of risks (Agbabian, et al, 1991; Harmathy, et al, 1989, Michalski,
1990b, Reilly, 1996). It is currently possible to estimate magnitudes of most risks to
within one order of magnitude of uncertainty. For those remaining risks, which cannot
be estimated reliably, there are four possibilities:
1) The risk can not be quantified accurately but it is known to be so high that it must be
accorded a high priority for mitigation.
2) The risk can not be quantified accurately but it is known to be so low that it can
be ignored.
3) The cost of mitigating the risk to a known, low level is less than the cost of research
or consultation needed to determine the magnitude of the risk accurately.
4) External consultation or new research is required to establish the magnitude of the
risk.

The risk assessment of the CMN collections resulted in identification of risks

belonging in each of these four groups. What is important to note is that a lack of
precise, concrete information does not cause the system to collapse but does result in
a recommended course of action.

2.2 RISK MANAGEMENT

After identifying and estimating the magnitudes of risks, potential means for
controlling risks must be identified and evaluated. To facilitate the comprehensive
identification of means of control, we have recognized three general methods of control:
1) Eliminate the source of the risk,
2) Establish a barrier between the source of the risk and the object/collection,
3) Act on the agent responsible for the risk.
Each of these methods of control might be implemented at one of seven possible levels
for control: 1) Location, 2) Site, 3) Building, 4) Room, 5) Storage unit, 6) Object, 7)
Policy / Procedure.

As was the case with identification of risks, having a framework within which
means of control can be identified greatly facilitates comprehensive identification of all
possible means of control.

At this stage, selected mitigation strategies are evaluated in terms of costs, risks
and benefits during both an implementation and a maintenance phase. Many strategies
will result in temporary increases of certain risks during the implementation phase,
especially if construction or extensive movement of the collection is entailed. Early
identification of these temporarily increased risks is beneficial. Consideration of all
possible benefits to the institution will often result in identification of benefits that are
not directly associated with collection preservation but that might be used to leverage
the project either in terms of cost or in corporate priorities.

Although the principal outcome of the risk management exercise now appears
as this set of well-defined projects given priorities according to their risk-cost-benefit,

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 3

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints
the process will result in many valuable decisions. In short, many specific risks are
dismissed during the magnitude of risks estimation stage as being insignificant (e.g.,
meteorite impact on collection), many can be mitigated to a great extent by very low
cost methods (e.g., implementing a system to recall overdue loans), and many can be
dealt with in groups as part of a major project (e.g., rehousing collections in a new
facility).

3. CATEGORIES OF SPECIMENS
For a preventive conservation program to succeed in eliminating unwanted
damage and loss in the most cost-effective manner possible the relative value of collec-
tions being protected must be considered. Because most collections will contain both
objects of very high and objects of very low value, it is not reasonable to compare
collections as being more or less valuable. Rather, individual objects within collections
must be accorded a value, at least a sense of worth to the institution.

This has been done by the Netherlands Ministry of Welfare, Health and Cultural
Affairs (1992 a, b) as part of their Delta plan for the Preservation of Cultural Heritage.
In this plan objects in museum collections were classified into four groups ranging
from: A) the most significant objects, through B) material considered as important in
a documentary sense, and C) the reserve or archival parts of collections to D) material
that could and perhaps should be deaccessioned (Cannon-Brookes, 1993).

For CMN collections, a system of categories of specimens (Price and Fitzgerald,

in press) has been established to group objects according to their value and importance.
These categories are summarized in Table 1.

Table 1. Descriptions of specimens assigned to each of five categories in the CMN

hierarchy of collection object value.

One of our primary institutional responsibilities is towards the preservation of

Category 1 specimens. Consequently, projects to mitigate risks to these specimens will
be accorded a high institutional priority. With the exception of Category 1 specimens,
due to grouping of collection materials, it is difficult to grade priorities for preventive
conservation functions and projects among all of the other categories. Exceptions to this
occur in some collections where Categories 1 and 2 and/or where Category 5 objects are
housed separately from the remainder of the collection.

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 4

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints
 Although some conservators might still object to this stratification of level of
care, it really just formalizes what we know to be true and already are doing. I am sure
that nobody would argue that the constitution of a country deserves a higher level of
care than the specifications for supplies of toilet paper, even if they are both held in the
same archives collection. Formal recognition of this difference in importance and value
of objects in collections is one of our most important tools for focusing resources for
preventive conservation to achieve the maximum cost-effectiveness in retaining
collection value over time.

4. COLLECTION PROFILING
Certain preventive conservation issues affect only limited numbers of objects
in collections. These issues are not well resolved by the broad perspective of a risk
assessment and require more detailed surveys. For large collections, doing object-by-
object surveys is not practicable so that some surveying strategy must be adopted.
Considerable work in the design of such surveys has been done in the United Kingdom
in recent years and is well summarized by Keene (1991).

In the field of natural sciences collection management, a system for profiling

both the conservation and collection management needs of a collection has been
developed for entomology collections (McGinley, 1993). This system uses the standard
entomology tray as the unit of measure and assigns a value between one and seven
according to the conservation needs and processing stage of each drawer. As the CMN
wanted to address both preventive conservation and collection management processing
issues in our survey procedure, the entomology system was used as a starting point.

In a 1994 workshop facilitated by Ron McGinley, CMN collection staff initiated

development of our own collection profiling system. The concept of a storage unit,
either a shelf, a tier of shelves, a cabinet, a drawer, or similar unit was retained. Instead
of a single numeric descriptor of the state of objects in the unit we have adopted two
descriptors, each ranging from one to four. It is interesting to note that the adoption
of four levels, instead of three or five, was the recommendation reached independently
by the United Kingdom Institute for Conservation (UKIC) working party on collection
condition surveys (Keene, 1991). One of the CMN descriptors applies to the level
of processing of objects, in a collection management sense, from initial preparation,
identification, and cataloging to final integration into the collection. Although this is of
equal importance to the preventive conservation descriptor in terms of setting collection
maintenance standards and priorities, it is not further discussed here as it is not part of
the primary focus of this paper.

There are two main purposes for the preventive conservation aspect of our
collection profiling system. These are:
1) To establish a basis for defining resource requirements for continuous maintenance.
That is, collections should not change level of care as a result of lack of resources for
continuous maintenance.

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 5

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints
2) To define, cost and prioritize remedial maintenance projects required to address spe-
cific preventive conservation issues in collections.

The working definitions of each of the four levels of our preventive conservation
descriptor is given in Table 2.

Table 2. The four levels of the preventive conservation descriptor, working definitions of
each, selected specific examples of observations in dry or in fluid-preserved collections
that would result in assignment of that level.

Considerable work is being done to ensure uniformity between characteristics

used to define levels for all of our diverse collections. This will result in our ability
to assign levels to collection units that are truly equivalent in terms of risk to objects
regardless of whether those objects are dry plants, fluid-preserved worms, books,
microscope slides, objects of art, or anything else.

5. MAKING IT WORK TOGETHER

The CMN is implementing a business planning process to replace our existing
practice of developing a complete new set of work plans each year. Our preventive con-
servation plans are contained within the Museum’s overall business plan; they address
one of the four goals of the Collection Division. Specifically, the goal “to manage and
preserve collections, by controlling risks and performing essential maintenance, so that
present and future generations can understand and appreciate our natural heritage”.

The plan includes two parts: the ongoing activity of continuous maintenance
and a series of remedial maintenance projects. Resources for continuous maintenance
are assigned to ensure that overall risk to collections is kept at an acceptable level and
that no collection material in categories one to four drops in level of care. Collection
risk assessments will be repeated every five years and we anticipate repeating collection
profiles every two to three years. Both of these will serve as the basis for performance
indicators.

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 6

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints
 Remedial maintenance projects are defined on the basis of risk management strate-
gies or on analysis of collection profiles. They are given priority according to catego-
ries of specimens involved, ratios of cost to risk reduction and other benefits, relation
to high priority corporate projects, availability of external resources, and other
factors.

6. SUMMARY AND CONCLUSION

The CMN has made the leap from planning preventive conservation on the
basis of implementing as many control measures as possible to planning on the basis
of minimizing overall risk to collections. The latter approach ensures the maximum
possible return on our investment in terms of maintaining collection value. There are
some difficulties associated with doing this in advance of having all of the technical
information we would like. Nevertheless, planning based on achieving the desired
impact, that is to maintain the value of collections in a cost-effective manner, carries
advantages that greatly outweigh the difficulties.

ACKNOWLEDGMENTS
I gratefully acknowledge the contributions and assistance of all of the Canadian
Museum of Nature Collection Division staff in completing the work outlined in this
paper.

REFERENCES
Agbabian, M.S., W.S. Ginell, S.F. Masri and R.L. Nigbor. 1991. Evaluation of earth-
quake damage mitigation methods for museum objects. Studies in Conservation. 36:
111-120.

Canadian Museum of Nature. 1994. Assessing and Managing Risks to Your Collections.
Brochure.

Cannon-Brookes, P. 1993. The ‘Delta Plan’ for the preservation of cultural heritage in
the Netherlands. Museum Management and Curatorship. 12: 303-307.

Harmathy, T.Z. et al. 1989. A decision logic for trading between fire safety measures.
Fire and Materials. 14: 1-10.

Keene, S. 1991. Audits of care: a framework for collections condition surveys. Storage.
Preprints for UKIC conference, Restoration ‘91. London, U.K. 6-16.

McGinley, R.J. 1993. Where’s the management in collections management: Planning

for the improved care, greater use, and growth of collections. In Current issues, initia-
tives, and future directions for the preservation and conservation of natural history col-
lections. ed. C.L. Rose et al. Consejeria de Educación y Cultura, Comunidad de Madrid,
Direccion General de Bellas Artes y Archivos, Ministerio de Cultura, Madrid. 309-338.

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 7

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints
Michalski, S. 1990a. An overall framework for preventive conservation and remedial
conservation. Preprints of the 9th Triennial Meeting of the International Council of
Museums Committee for Conservation. Dresden. 589-591.

Michalski, S. 1990b. Towards specific lighting guidelines. Preprints of the 9th Trien-
nial Meeting of the International Council of Museums Committee for Conservation.
Dresden.583-588.

Michalski, S. 1994. A systematic approach to preservation: Description and integration

with other museum activities. Preprints of the 15th International Congress. Interna-
tional Institute for Conservation of Historic and Artistic Works. 8-11.

Netherlands Ministry of Welfare, Health and Cultural Affairs 1992a. Deltaplan, Preser-
vation of cultural heritage in the Netherlands.

Netherlands Ministry of Welfare, Health and Cultural Affairs 1992b. Deltaplan for the
preservation of cultural heritage in the Netherlands. Fact Sheet C-11-E-1992.

Price, J.C. and G.R. Fitzgerald. in press. Categories of specimens: A collection manage-
ment tool. Collection Forum._

Reilly, J.M. 1996. IPI’s Environmental Assessment Technology.This volume.

Waller, R. 1994. Risk management applied to preventive conservation. Preprints of the

15th International Congress. International Institute for Conservation of Historic and
Artistic Works. 12-16.

Waller, R. 1996. Risk Management Applied To Preventive Conservation. In: Storage of

Natural History Collections: A Preventive Conservation Approach. ed. C.L. Rose et al.
Society for the Preservation of Natural History Collections,
21-28.

Wolf, S.J. 1993. Conservation assessments and long-range planning. In Current issues,
initiatives, and future directions for the preservation and conservation of natural his-
tory collections. ed. C.L. Rose et al. Consejeria de Educación y Cultura, Comunidad de
Madrid, Dirección General de Bellas Artes y Archivos, Ministerio de Cultura, Madrid.
289-307.

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 8

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints
ROBERT WALLER received a BSc from the University of Manitoba and is accredited
by the Canadian Association of Professional Conservators (CAPC). After fifteen years
of work on mineral conservation he served as Head of Conservation for two years,
for the past three years he has been Managing Director, Collection Division at
the Canadian Museum of Nature. He has taught and consulted on preventive conserva-
tion issues throughout North America, Europe and the Caribbean. Address: Canadian
Museum of Nature, P.O. Box 3443, Station “D”, Ottawa, Ontario KIP 6P4, CANADA,
[email protected]

PREVENTIVE CONSERVATION PLANNING FOR LARGE AND DIVERSE COLLECTIONS 9

Robert Waller. Submitted to: American Institute for Conservation 1996 Presession Preprints