[Title page]
From the unsentimental viewpoint of a physicist, museum collections are stuff that is, physically, past its best, but has by chance or sentiment avoided being thrown away.
Many good scientists have studied the degradation pathways of materials, but given the infinite variety of delicacy of actual objects in collections there is something to be said for concentrating on optimising the container: the building which surrounds the collection and forms its primary shield against the weather.
This is my market town, Totnes, in south west England. Its houses use the best method yet devised for keeping walls dry in a windy and wet climate - slate hung vertically as well as used for roofing.But these are from the sixteenth century.
[Royal Ontario Museum]
For decades, the museum community has been shielded from researching the merits of vernacular architecture by inflexible environmental standards. The decades long obsession with 'best available technology' has produced buildings which simply have to be air conditioned to fulfill the stringent requirement for climatic constancy. So it has not been necessary, or competitive, to make buildings that function in themselves as climate shields against the weather. Quite harsh winter weather in the case of the Royal Ontario Museum.
[ROM interior]
To keep to the 20th century standard, the Royal Ontario Museum has to pipe conditioned air to the showcases because the building cannot provide the necessary, or rather the specified, climate.
[Queen Victoria's bathing machine]
Now we can celebrate the relaxation of standards for conservation climate, as shown in this recently constructed canopy for Queen Victoria's bathing machine at Osborne House on the Isle of Wight. Many of us will also celebrate the relaxation in bathing protocol which has made this item an amusing period piece.
[Flaking wall painting]
But have the new standards relaxed enough from the period pieces which were their predecessors? Or by an amount which allows real change in the design of museums, their stores and archives? I won't discuss if the standard is safe, though there is much debate about this. But I will comment on the limits it sets, in the context of the laws of nature, and the facts of the weather, and the properties of materials.
I take as my theme the British and European quasi-standard for archives PD5454:2012. The now obsolete BS5454:2000 has been enormously influential in the museum world, worldwide, because it was firmly, though ambiguously, numerical, in a way that appealed to contract lawyers for building projects, and to engineers commissioned to install elaborate air conditioning.
Now the stringency has disappeared. Even the document title - 'published document' rather than 'British Standard', reveals a lack of confidence in the product. And archivists everywhere are left feeling foolish that they loyally and proudly served a fierce standard, which is now regarded by anonymous authorities as unnecessarily strict.
[Ribe temperature]
But now it's time to talk about the weather. Here is the temperature record for Ribe on the west coast of Denmark. A typical northern European weather. It consists of a series of daily cycle, shown expanded in the inset, superimposed on an annual cycle.
It is a trivial matter to entirely suppress the daily cycle in a museum store or archive, even in a museum gallery.
[Quote 6.3.1 from PD5454]
PD5454 is quick with advice. It advocates thermal inertia to ensure stability.
[Temperature evolution through a brick wall]
Here is how heat capacity in a wall works. As heat from the daytime solar radiation is absorbed at the surface, it passes into the wall but is absorbed as molecular vibrations in the substance of the wall, then re-emitted so that the heat wave from the daytime solar radiation is delayed and diminished on passage through the wall. The perspective section through the wall on the right, shows how the temperature cycle within the wall diminishes as one approaches the inside surface. The inside temperature peaks about 12 hours after the outside maximum.
[Koeln archive ruin]
One disadvantage of such a heavily built construction was tragically revealed by the 2009 collapse of the Cologne City Archive. The massive rubble of the walls pinned down the crumpled archive documents so it took more than a year to rescue them.
[temperature through insulation]
There is another way to moderate the daily temperature cycle: insulate the wall instead of loading it with heat absorbing mass. Here is the performance of an insulated wall whose thickness has been adjusted down to give exactly the same temperature variation at the inside surface as the brick wall. So a deeper analysis of the building physics suggests that a lightweight, insulated archive is both cheaper and equally effective, and less burdensome when disaster strikes.
[Underground temperature]
However, PPD5454 is not entirely wrong in its advice, just imprecise. thermal mass is capable of moderating the annual temperature cycle. Walls would have to be about four metres thick to achieve this degree of buffering, but there is plenty of heat absorbent material under the building, so an uninsulated floor laid directly on the ground will greatly reduce the annual temperature cycle within the building above it.
[Temperature in Ribe store]
Using the ground as a heat sink reduces the annual temperature cycle within the building to around 8 K. (The kelvin temperature is customarily used in physics to indicate temperature difference when it could be confused with the actual temperature)
This swing from 16C to 8C is very modest, and being below human comfort temperature, it contributes to preservation of the collection by reducing the chemical decay rate.
Notice also that even though the heat sink is at the bottom of the building, even in summer the temperature difference between the floor and the six metre high ceiling is less than two degrees.
[PD5454 quote 2]
So how does this temperature cycle fit with PD5454? Not well. Although purely paper stores are allowed to sink to 5C, a general collection is limited to 13C minimum. The reason given is the evidence from a single published paper that beeswax seals precipitate natural components below this temperature. This is a good case for risk analysis enthusiasts to chew on: the welfare of a few wax seals that have never before been below 13C against the benefit to most other unstable materials from a lower average temperature.
The bottom line of this quote introduces the other major preoccupation of conservators - the relative humidity. Can building physics contribute to humidity stability. It turns out that the permitted range from 35 to 60% is unnecessarily generous. In archives and stores we can do much better than that.
[Suffolk record office climate]
The relative humidity within the Suffolk record office stays within a 5% span all the year round. There is no explicit humidity control. This stability is achieved by winter heating alone, to a fixed 14C, which conveniently brings it in line with PD5454 temperature requirements. This is achieved by buffering by the abundant paper documents, which release water vapour during the winter, so that the RH declines by only a few percentage points. In the summer, the inside temperature is often below the outside temperature, risking a high inside RH, but the paper absorbs moisture from the infiltrating air so the RH only rises a few percentage points. The green trace shows the magnitude of the imbalance between inside and outside water vapour in the air.
At this point I will anticipate a frequently voiced objection to humidity control by irreplaceable national treasures. The water molecules sorbed and desorbed by paper at moderate RH are very loosely bound, with a very short residence time before they hop off and are replaced from the vapour in the surrounding space. The object does not own its water molecules - they are rapidly exchanged with the environment, regardless of how that environment is controlled.
Sadly, the temperature span in the Suffolk record office exceeded the range permitted by the previous BS5454:2000, so it was air conditioned, a year before the successor document permitted its temperature cycle as OK for anything. Both thes documents describe themselves as advisory, but many decision makers lack the confidence to boldly go where the British Standards Institution advises against.
[cotton isotherms] Optional slide - depends on the time available
The influence of temperature change on relative humidity is worth a comment at this point, since I am advocating allowing a temperature cycle yet claiming that humidity constancy is easily achieved. This graph shows that the balance between absorbed water in cotton and the ambient RH hardly changes with temperature. In a buffered environment the RH falls slightly with falling temperature, at about 0.3% RH per degree. In an unbuffered space the RH rises with falling temperature, at about 3% per degree. The break even point where there is no change in RH with temperature is attained at about 100g/m3 of cotton, which is less than in an archive. So in practice an entirely sealed archive will see a fall in RH of around seven percentage points on moving between the permitted limits for paper from 25C to 5C. This is not a bad thing, on the contrary it indicates that the paper keeps a constant water content and therefore nearly constant dimensions.
[bricks in corner]
Many museum objects are not moisture absorbent but nevertheless susceptible to damage from RH extremes. It is possible to augment the buffer capacity of a museum store with low air infiltration by lining the walls with unfired clay bricks. This is the cheapest available humidity buffer. readily snatched from the brick works production line just before entering the final firing.
[climate in the clay room]
It is effective, with a performance quantified in an article by Padfield and Jensen listed in the more comprehensive article on the website. Their simulated prediction is the red line. The spiky real climate record is caused by sunlight entering the room and temporarily overwhelming the buffer capacity of the clay lining the walls.
One can conclude that designing a museum store is quite simple: it should be a lightweight superstructure on an uninsulated floor, with winter heating to reduce the annual average RH and accord with PD5454, or summer dehumidification to keep the annual average temperature lower and thus minimise the rate of chemical decay. Economically, dehumidification of a nearly airtight building with limited movement of objects and people is very cheap. It is only needed in summer when solar energy is abundant.
[Birkehoej section]
I finish by pointing out that before the twentieth century abandonment of designing buildings for their purpose, there was a fine tradition stretching back at least 5000 years of designing for durable storage of inanimate but valued relics. This is a cross section of a grave mound excavated in Denmark, showing the subtlety of its construction, designed to keep the interior dry and the temperature even. Its two metre thick capping puts it firmly in agreement with PD5454, but the only available lightweight insulation was sheeps wool and maybe the builders were not confident of its durability over thousands of years.
[Acknowledgements]
I thank my colleagues in the museum climate group at the National Museum of Denmark, and others who have contributed data and drawings.
This lecture, and an associated, more formal article, are free to download from my website.
Lecture: http://www.conservationphysics.org/cpw/uploads/Storage/padfield-parislecture-v01.pdf
Article: http://www.conservationphysics.org/cpw/uploads/Storage/building-phys-for-museum-stores.pdf