Tobit Curteis and Naomi Luxford
|The effects of moisture: left, pitting caused by the dissolution of soluble components (Photo: Tobit Curteis Associates) and, right, microbiological growth
(Photo: Holy Well Glass)
Despite its fragile and brittle nature,
stained glass has survived in many
churches and cathedrals from as early
as the 12th century. Often it is the last vestige
of the rich colourful decoration that covered
walls, fittings and furnishings in the medieval
church interior, and early examples are of
huge significance. Unlike other works of art,
stained glass also forms an integral part of
the structure and envelope of the building.
As such, it is of particular importance that
stained glass of all periods should be conserved
in situ wherever possible, an approach which
often requires the use of protective glazing.
As part of an initiative to improve the
understanding of protective glazing systems,
a detailed research project commissioned by
the Building Conservation and Research Team
at English Heritage is currently under way.
Seen from a distance inside a church, historic
stained glass can look remarkably well preserved.
However, distance and transmitted light often
disguise problems which are only apparent
on much closer inspection. Medieval stained
glass, and some later painted decoration, are
often chemically unstable as a result of the
materials and techniques used to create them. In
particular, the leaching of potassium and sodium
ions from the glass in the presence of water can
cause severe physical deterioration of the glass
body. Stained glass windows are also subjected
to wind-loading which can damage the entire
structure. The risk of vandalism and deliberate
damage add further weight to the arguments
for protecting the exterior face in particular.
In addition, the internal face can
be damaged by repeated high levels of
condensation, which facilitate the dissolution
of soluble compounds within the glass paint
and body causing pitting. Differential thermal
stresses can lead to delamination of paint and
enamel layers. High levels of condensation
can also lead to substantial microbiological
growth on the surface of the glass. Not only
can this growth be disfiguring but it can
also retain moisture and the physical and
chemical side-effects of the lifecycle of the
organisms can cause further damage.
OPTIONS FOR PROTECTION
Although there are a number of ways of
protecting stained glass from physical
damage and vandalism, for example the
use of wire mesh, the only system which
can control chemical and environmental
deterioration in situ is protective glazing.
Protective glazing has a long history:
it was used at York Minster from 1856 and
Lindena church in Germany from as early as
1796. However, the benefits of this approach
were first recorded during the 1960s in Bern
Minster in Switzerland, where some of the
stained glass had been reinstalled in frames
behind new glazing after the war. These
were in noticeably better condition than
the panels that had been reinstalled in their
original positions, which were unprotected.
Protective glazing is now widely used to
conserve important and vulnerable stained
glass. However, just as no two windows or
churches are the same, no two protective
glazing systems are the same. As a result
it has been difficult to determine the best
design features for specific installations.
Protective glazing involves the installation of a
layer of new glass on the exterior of the window.
In some cases the historic glass is left in situ,
while in others it is moved forward (towards the
interior) on a metal support with the new glazing
installed in the original grooves in the tracery.
In most cases the gap between the original
glass and the protective glazing is ventilated at
the top and bottom, which allows air to pass
through the interspace between the two. Some
systems are vented to the outside (externally
ventilated), some to the inside (internally
ventilated) and some combine the two.
In some cases sealed units that are more
similar to double glazing have been used.
However, due to the difficulty in creating
an effective seal, significant condensation
can occur and this approach is now rare.
Most current protective glazing is internally
ventilated, allowing the historic glass to
be surrounded by air from within the
building, which also, generally, has the
advantage of lower levels of pollution.
||New protective glazing in the south oculus window
at Canterbury Cathedral, partially obscured by the
heavy structural armature and grille
||Poorly designed and badly maintained grilles obscure
the exterior of a stained glass window, and (below)
shadows cast by a grille disfigure the image inside
(Both photos: Tobit Curteis Associates)
One of the main concerns with the use of
protective glazing is its impact on the appearance
of the building. A wide variety of approaches
can be considered to modify the external
appearance of the protective glazing, including
glass type, glass surface treatment, and whether
the glazing is constructed in panels or is leaded
to imitate the original glass and mounting.
However, change in appearance is nothing
new. Corroding 20th-century metal mesh is
often tolerated because it appears always to have
been there. Increasing opacity and distortion
of the original glass is tolerated because it
occurs gradually. However, both cause at least
the same level of visual impact as protective
glazing when viewed externally. Internally,
the shadows cast by grilles are particularly
damaging to the appreciation of works of art.
Over the past 40 years research has been
carried out, mostly in France and Germany,
into the efficacy of different protective glazing
systems. Between 2002 and 2005 an EU-funded
project (VIDRIO – see Recommended Reading) studied protective glazing
in a number of churches and cathedrals across
Europe. This project looked at how windows
were ventilated, the amount and duration of
condensation periods, pollution levels within the interspace, and temperature changes.
There was a wide variation in the windows
tested in terms of the geometry of the window,
the size of the air vents and the depth of the
interspace. VIDRIO reported that the size
of the interspace and dimensions of the air
vent were fundamental elements for the
ventilation of protective glazing. However, there
were no firm guidelines on what sizes were
critical when designing protective glazing.
The VIDRIO research also showed that glass
corrosion is accelerated by prolonged exposure
to liquid moisture from condensation, as well as
by high levels of gaseous pollution. For unstable
glasses these effects are increased as the glass
develops cracks allowing pollutants and moisture
to continue to penetrate beneath the surface.
ENGLISH HERITAGE RESEARCH
Because of the limited research data available
on specific design details, environmental
effects and the aesthetic impact of protective
glazing, there is considerable difficulty for
conservators, advisory bodies and clients in
evaluating the benefits or drawbacks for specific
installations, in particular in parish churches.
In order to provide the necessary
information to allow informed discussions
to take place, in 2012 English Heritage
commissioned a research programme to look
at both the technical and aesthetic issues
involved and, in particular, questions relating
to energy efficiency. This included a detailed
literature study to establish the current state
of understanding (including translation of
key materials), computational fluid dynamic
modelling to assess how specific design
details affect functionality, and environmental
monitoring and evaluation of existing and
new protective glazing systems, to provide a
better understanding of actual installations.
The research programme is expected to
be completed in 2015 with the results published both in academic journals and as
an English Heritage advisory document.
|The ventilation gap at the base of an internally ventilated system: in this example the stained glass has been left in the original glazing grooves and the base of the glass tilted
forward to allow airflow, with a similar gap at the top. (Photo: Holy Well Glass)
The preliminary results of the research
have demonstrated that protective glazing
provides significant benefits to vulnerable
stained glass. In almost all cases it provides
protection against physical damage, windloading
and direct rainfall.
For internally ventilated systems, the
range of temperature fluctuations can be
significantly reduced due to the thermal
buffering provided by the protective glazing.
The lowest temperature values on the
historic glass are generally maintained at a
higher level than on the unprotected glass;
this reduces the dew point temperature and
so minimises condensation. In most welldesigned
installations, condensation is entirely
prevented on the historic glass. Condensation
continues to occur on the modern glass but the flow of air through the interspace ensures
that any condensation rapidly evaporates.
The reduction in condensation on the
internal face of the historic glass reduces
deterioration associated with both dissolution
of the soluble components of the glass and
failure of the paint and enamel layers. In
addition it has a direct and significant effect
in reducing microbiological growth.
Research on grisaille glass has shown
that reducing the range of rapid temperature
change is particularly important for glass
with applied layers of paint or enamel (in
grisaille work, designs are generally painted
onto the face of predominantly ‘white’ glass
in blacks and greys). In these types of glass,
large and rapid temperature changes can
lead to stresses developing at the boundary
between the applied layers and the body
glass, resulting in delamination and flaking.
Many of these benefits were also observed
on externally ventilated systems, although they
were found to be less efficient, providing a
lower level of protection. Externally ventilated
systems have the additional disadvantage that
they can allow wind-driven rain to enter the
interspace through the ventilation openings.
One much-repeated concern regarding
protective glazing has been that it reduces
rainwater washing of the historic glass
surface, leading to the build up of dirt and
pollutants, including particles that retain water.
While surface washing will be prevented,
protective glazing has been shown to lead
to lower levels of particulates and gaseous
pollution. This indicates that the loss of the
washing effect is unlikely to be critical.
Data collected from many UK sites has
demonstrated that even an inefficient system
of secondary glazing can offer significant
protection, minimising the impact of
external factors such as wind and rain and
greatly reducing internal condensation.
||Thermographic image showing heat loss though a
stained glass window and the thermal buffering effect
of the protective glazing fitted to the upper centre light
(Photo: Tobit Curteis Associates)
||Computational fluid dynamics modelling to assess
how design options affect air flow and performance
(Photo: Element Energy)
With the aid of computational fluid dynamics,
a computer simulation of protective glazing
systems has been created to identify the
key thermal performance features and
phenomena. This has shown that the air vents
should ideally be placed at the very top and
bottom of the glazing to maximise airflow.
However, air vents on the front face can
still be effective if they are properly sized.
To promote optimal airflow the air vents at
the top and bottom should also be equally sized
and distributed as evenly as possible across
the width of the glazing. While larger vents
improve interspace airflow – and therefore
thermal buffering – the improvement slows
when the ratio of vent area to interspace area
(measured horizontally) exceeds 30 per cent
as turbulence increases. The use of fine wire
mesh and other materials covering vents also
decreases efficiency of airflow so an unrestricted
vent size of 25-30 per cent is desirable.
Further modelling of realistic geometries
for stained glass windows with protective
glazing, such as small and large lancet
windows and tracery lights, is planned. This
should also further test different air vent
geometries and different interspace depths.
Previous research into the energy efficiency
benefits of ventilated protective glazing has
been limited. Studies in the USA on the use of
storm windows with clear glass have shown a
significant reduction in heat loss, even when
the units have some air infiltration. While these
systems do not achieve the same levels of heat-
loss reduction as sealed secondary glazing or
double glazing units, an increase in efficiency
of up to 80 per cent is commonly observed.
The use of low-emissivity glass can also have
a significant effect in reducing heat loss.
However, the amount of glazing in a church
or cathedral relative to the amount of masonry,
is likely to remain the critical parameter in
terms of how much heat is lost via the envelope
and therefore the extent to which protective
glazing can improve overall thermal efficiency.
Nevertheless, in a church with extensive
glazing, the impact could be significant.
The extensive environmental monitoring
data from protective glazing systems in a
number of UK case study sites confirmed
previous published findings and provided a
more detailed understanding of how different
systems function. The research outcomes
will draw together these findings with the results from previous studies in Europe and
will provide detailed information on the
benefits of protective glazing. Together with
the mathematical modelling, the analysis
will enable an overview of the critical design
features for protective glazing systems
and will help to identify which features of
the design are most significant in building
efficient systems. The aesthetic impact
and functional performance of protective
glazing can then be evaluated against
the deterioration of the stained glass.
The information will be available as a
guidance document from English Heritage
for conservation practitioners and advisors as
well as those charged with the care of historic
D Anderson, ‘Stained Glass and Its Decay’,
The Conservation and Repair of
Ecclesiastical Buildings, Cathedral
Communications Limited, Tisbury, 1996
M Bambrough, ‘Aesthetic Protective
Glazing’, Historic Churches, Cathedral
Communications Limited, Tisbury,
F Becherini et al, ‘Thermal Stress as a Possible
Cause of Paintwork Loss in Medieval
Stained Glass Windows’, Studies in
Conservation, 53, 2008
A Bernardi et al, ‘Conservation of Stained Glass
Windows with Protective Glazing’, Journal
of Cultural Heritage, 14, 2013
Corpus Vitrearum Medii Aevi: Medieval
Stained Glass in Great Britain,
RHM Godoi et al, ‘The Shielding Effect of the
Protective Glazing of Historical Stained
Glass Windows’, Atmospheric Environment,
T Husband et al (eds), The Art of Collaboration:
Stained Glass Conservation in the 21st
Century, Harvey Miller, Washington, 2010