Jean Tétreault

Canadian Conservation Institute

STANDARDS FOR LEVELS OF POLLUTANTS IN MUSEUMS: PART II

Introduction

At the first meeting of the Indoor Air Pollution Working Group in Glasgow in 1998, specifications for indoor pollutants in museums have been proposed (Tétreault 1998). There was agreement to replace the terminology "use best control technology" by maximum levels of some pollutants. As the most simple approach, a maximum concentration for a specific pollutant based on its unpolluted outdoor levels was proposed. Even the group admitted the need for more explicit specifications, but the approach to define suitable maximum concentrations required further consideration.

Approaches
At the Amsterdam 1999 meeting, five approaches to establish the maximum pollutant levels have been considered. All approaches have limitations in their application:

Acceptable Risk Concentration: concentration below which the risk for damage is minimal. The risk is based on accumulated experimental data (NOAEL).

Background Level: level based on average levels from unpolluted outdoor environment (clean troposphere). This simple approach was used in conservation for the last 20 years (Thomson 1986). It has been observed that objects stored far from industrial or urban environment remain in much better conditions over many decades or centuries than those exposed to polluted urban environment.

Dosage: maximum cumulative flux (concentration x time) allowed for a pollutant (Brimblecombe 1998). This approach is used for pigments and dyes light fading. It assumes a linear pollutant concentration dependency with the object. Contrary to light, this dependancy may not be always observed with pollutant. For a same dosage a high concentration of gas exposed to a short time will usually not give the same amount of deterioration than a low concentration at a longer exposure. The corrosion of lead by acetic acid does not seem to follow a linear relation (Tétreault et al. 1998). This approach has not been widely used in the indoor pollution field. Many experiments on the concentration dependency are needed to establish one specific dosage.

No-Observed Adverse Effect Level (NOAEL): level at which damage is not observed for a specific setup (analytical method, exposure time, temperature and presence of other reactants or catalysts). The information found can be very limited depending of the experimental setup and can bring premature conclusion on the concentration dependency of the object. The proper usage of the term would include specifications of all parameters necessary to make the value meaningful. For example, a useful syntax for NOAEL would include: NOAEL = Concentration (temperature, relative humidity, time, property measured).

Threshold: level at which reaction cannot happen in any time. It relies on reaction kinetics and thermodynamics.

For example, a theoretical approach using the Gibbs energy gives the critical concentration of an equilibrium chemical equation (Raychaudhuri, 2000). The effect of environment parameters such as the presence of catalysts and the history of the object complicate the calculation of the critical concentration. Little work on thresholds has been done in the conservation field.

After discussion, the working group agreed to investigate the feasibility of an approach on concentration dependency such as the dosage (Low-Observed Adverse Effect Level (LOAEL) x Time) approach. For the most accurate dosages, it was recommended that the environmental conditions should be close to those found in museums. The dosage can normalize on one year exposure time (µg/m³·y or ppb·y) for an easy conversion of the dosage in decades or centuries. The determination of a dosage for a specific object-pollutant combination should consider also the presence of other elements involved in the interaction (see Appendix 1). In the field of materials testing, it is interesting to notice the use of empirical damage acceptable levels to estimate the useful lifetime such as the 0.1 degree of yellowing for adhesives (Down et al. 1996) or a specific amount of lost of tensile strength. For some objects, this approach can be considered. The two following examples show the potential of the dosage approach:

Lead and acetic acid:
The dosage at which no damage was observed for lead exposed to acetic acid at 54%RH and at room temperature was found to be 220 µg/m³ · y (Tétreault et al. 1998). A concentration of 220 µg/m³ of acid is the dosage for the period of one year. This dosage does not guaranty the absence of corrosion on lead for many decades. However, several years of monitoring at the British Museum, showed no corrosion on lead objects when the acetic acid levels in the enclosure were less than 317 µg/m³ (Thickett et al. 1998). With the dosage approach (assuming linear concentration dependency and controlled presence of the other elements involved in the reaction), the level of acetic acid to avoid corrosion for a period of 100 years should be less than 2.2 µg/m³. This value is in the order of magnitude of the acetic acid background level.

Silver and hydrogen sulphide:
It was observed that silver objects have to be cleaned after six month exposures to hydrogen sulphide at 570 ng/m³ (Watts 1999). If we assume that the LOAEL for six months is around 400 ng/m³, the dosage can be extrapolated to 200 ng/m³·y. A concentration of 200 ng/m³ is close to the background level of the hydrogen sulphide. To avoid the silver from tarnishing for 100 years, the concentration of hydrogen sulphide will have to be kept at a challenging, low level of 2 ng/m³. Materials that are very sensitive to a certain pollutant, such as silver to hydrogen sulphide, can not be used to define a general reference dosage, they need to be dealt with separately.

Pollutants
Sulphur dioxide, nitrogen oxide(s), ozone, acetic acid, formaldehyde, hydrochloric acid and fine particles are the common pollutants referred to in the museum standard literature. With the exception of acetic acid and formaldehyde, maximum levels have been defined for all the pollutants, based on the background level or on the best HVAC system technology. The same dosage approach can be considered for acetic acid and formaldehyde as well. Significant differences between two approaches can lead to further examination of the values so that the most appropriate maximum levels can be defined.

Formaldehyde as a pollutant raises an interesting issue. As such, formaldehyde does not really corrode metal but its oxidized form, formic acid, does. The presence of peroxide will lead to a fast oxidation of formaldehyde (Raychaudhuri 1999). Formaldehyde will remain included in the list of pollutants because of its known potential to become a truly damaging compound.

At the meeting, there was also an interest to include ammonia, formic acid and hydrogen sulphide. During the revision of the document, it has been suggested to replace hydrochloric acid by nitric acid since there is more data available on the interaction of the latter with objects. The list of pollutants remains open to reconsideration depending on the knowledge of the pollutant concentration dependency for objects.

Presentation Format
Table 1 presents the preliminary specifications format. The working group was in favour of the presentation of the specifications with different classes of control and on the description of collection risks and benefits in a similar way as the temperature and relative humidity specifications used in the Chapter on Museums, Libraries and Archives from the 1999 edition of the American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (ASHRAE 1999). The class of control allows some ranges of freedom: not all the collection, enclosures or rooms require the same control level. It is more suitable and feasible to have a strict control for storage areas with limited access than temporary exhibition areas. The preliminary proposition on class of control is based on three levels ABC where A being the "minimal risk" pollutant concentration for a period of 100 years for the most sensitive object; B and C are for a period of 10 years and 1 year respectively. As seen with lead, the most sensitive to acetic acid vapor, the maximum level of acetic acid for general museums of each class of control could be 2.2, 22 and 220 µg/m³ for the class of control A, B and C.

The maximum pollutant concentration for each class of control should be applied in the building levels as well as in the enclosure levels. It is known that the level of pollutants is usually higher in enclosures than in buildings due to the lack of ventilation in enclosures. The rationale of each pollutant maximum should be well documented. The feasibility or the impact of the specifications should also be investigated.

Communication
The development of the specifications will be done by consulting members of the Indoor Air Pollutant Working Group. Other groups of interest dealing with the environment, the air quality or life expectancy of materials will be consulted for sharing experiences. Curators should also give their input by defining what is unacceptable damage for different types of objects. This information will help to determine on which experimental design the laboratory studies into NOAEL values should be based. For example, what is the most unacceptable sign of deterioration of paper exposed to pollutants over time? Is it embrittling, yellowing, acidity, the decrease of the degree of polymerisation or a mixture of them?

Conclusion
This long term project will help to update or refine the existing specifications for pollutants in museum, libraries and archives. These specifications are primary tools in the prevention of damage caused by pollutants. The list of pollutants to control will be revised and the approach of concentration dependency will be seriously considered to establish the maximum pollutant concentration for different class of control. Resources will be required to investigated object-pollutant interactions to establish the most accurate dosages. Simple analytical methods have to be optimised for the range of pollutant levels covered in the different class of control.

Those who are interested in this initiative and want to give feedback on the approach taken, provide information such as well documented concentration dependency for different materials, provide the rationale for existing specifications in museums indoor air quality or, provide "life expectancy" concepts (like those used in Archives) in other fields, are welcome to contact the author.

Appendix 1.
Damage rate depends on: nature of the object (including history and type), main pollutant level (primary reactant(s); possible synergy), other reactants levels (secondary reactants; including oxygen, water vapor and may be some oxidants) and catalyst levels (metallic particles), deposition velocity, temperature and radiation.

Total damage: damage rate x exposure time.

References

Table 1:

Indoor Air Quality Specifications for Museum, Gallery, Library, and Archive Collections

		Maximum Levels in Controlled Spaces for Different Class of Control, µg/m3; (except for relative humidity)				Reference Concentrations, µg/m3
		A	B	C		Clean Troposphere	Polluted Urban Air	Polluted Indoor Air
Acetic acid		-	-	-		-	-	-
Ammonia		-	-	-		-	-	-
Formaldehyde		-	-	-		-	-	-
Formic acid		-	-	-		-	-	-
Hydrogen sulfide		-	-	-		-	-	-
Nitric acid		-	-	-		-	-	-
Nitrogen dioxide		-	-	-		-	-	-
Ozone		-	-	-		-	-	-
Sulfur dioxide		-	-	-		-	-	-
Relative humiditya		%	%	%
Particle mattersb		-	-	-		-	-	-
Collection Risks and Benefits		No or small risk to sensitive objects...	Moderate risk to sensitive objects...	High risk to sensitive objects... Damage may occur to sensitive objects over a year period...		No or small risk to sensitive objects. Except for ...	High risk to sensitive objects... Damage may occur to sensitive objects over a year period...

a : or historical annual average for permanent; Temperature set between 15 and 25°C

b : Particles below 10 µm diameter