Historic paint analysis and specifying redecoration

Historic paint analysis establishes the sequence of decorative schemes on a surface and, where the evidence survives, the composition of the earliest of them. What it does not establish is what the room looked like. Pigments fade, oil media yellow, varnishes darken, and a cross section read under a microscope gives you a layer, not an appearance. Any redecoration specification built on analysis has to carry that distinction through to the paint schedule, or the finished room will be an honest match to a degraded sample.

For a reviewer, the question is narrower than "is this a good report". It is whether the sampling supports the conclusions and whether the specification that follows is compatible with the substrate.

What analysis can and cannot establish

Cross section microscopy on a resin mounted sample shows the full build up from substrate to present surface. Polarised light microscopy identifies pigment particles. Elemental and molecular techniques identify inorganic pigments and binding media, and ultraviolet fluorescence separates varnish layers from paint layers.

From that you can usually establish three things: how many separate decorative schemes a surface has carried, the composition of the earliest surviving scheme, and an earliest possible date for any given layer from the pigments present in it. Pigment introduction dates are the most useful of these. Prussian blue is not available before the early eighteenth century, zinc white becomes common from the mid nineteenth, and titanium dioxide white is a twentieth century material, so a layer containing it cannot be Georgian however convincing its colour looks under the scalpel.

What analysis cannot do is recover a lost distemper scheme. Soft distemper was routinely washed off before redecoration, so plaster surfaces frequently show an early oil scheme on joinery and nothing at all from the same date on the walls. The absence of a layer is not evidence that the wall was undecorated, and a report that treats it that way should be sent back.

Where to take samples and how many

The value of the exercise is decided before the microscope is switched on. Sampling has to target survival, and survival concentrates in places later decorators could not reach.

Look behind shutters and shutter boxes, in the rebates of sash boxes, under later architrave, behind pipe runs and radiators, on the returns of cornice mouldings, inside cupboards and on the backs of doors. Each of those is a place where a scraper stopped.

Sample each element separately. Skirtings, dados, joinery, plaster, cornice and ironwork were painted in different schemes at different dates, and one sample from a door will not describe the room. Six to twelve samples for a single principal room is a normal range for a scheme where the decorative history matters, fewer where the question is simply whether an original scheme survives at all.

Every sample needs a number, a location marked on a measured drawing or a rectified photograph, a photograph of the sampling point before and after, and a note of the substrate. Samples taken without that record are of limited evidential use, because nobody can later say which part of which element they came from. The same discipline applies as in recording a building before demolition or conversion: the record is the deliverable, not the photographs.

Sampling itself is minor physical intervention on a listed building. On a heavily significant interior the sampling locations are worth agreeing with the conservation officer in advance, and on rare occasions they form part of a consent.

Reading a stratigraphy

A layer count is not a scheme count. Each decoration usually leaves several layers: a primer or size, one or two undercoats and a finish, sometimes with a varnish or glaze on top. The analyst's job is to group those into schemes and number them from the substrate upwards.

Two habits separate a useful stratigraphy from a list. The first is grouping with reasons given, so that the reader can see why layers four to six are read as one scheme rather than three. The second is consistency of numbering between samples, so that scheme three on the skirting is the same campaign as scheme three on the door.

Watch for discontinuities. A surface that shows fifteen schemes next to one that shows five usually means the second has been stripped, and stripping is itself part of the building's history worth recording. Where a moulding shows no early layers at all, it is probably a replacement, which is useful information for a Heritage Statement whether or not the redecoration goes ahead.

An illustrative example. A first floor drawing room in a late Georgian terrace shows, on the shutter reveal, a lead based stone colour directly on the primer, then a deeper drab, then a run of pale schemes from the mid nineteenth century onward. The cornice, sampled separately, shows distemper traces only from the 1870s. The honest conclusion is that the earliest surviving joinery scheme is a stone colour, and that the original wall treatment is not recoverable from the fabric.

Matching a colour without misreading it

A matched colour has to be stated in a way that survives a change of supplier. Munsell notation, or a measured colour reference taken from the exposed and cleaned layer under controlled lighting, both do that. A manufacturer's colour name does not, because ranges change.

Then comes the interpretation. Linseed oil yellows, so a sample that reads warm now may have been cooler when applied. Some pigments, particularly early organic reds and some greens, fade. Lead white in an oil medium darkens under damp conditions. A specification that adopts the measured value without any adjustment produces a room that is faithful to the sample's present condition, which is a defensible choice only if it is a stated choice.

Sheen level and binder matter as much as hue. A flat oil finish and a modern eggshell in the same colour read as different colours in raking light, and the difference is visible in every photograph of the completed work. State the sheen, state the binder, and require sample panels on the actual substrate for approval before the main work.

Breathability and substrate compatibility

This is where specifications fail in service rather than on paper. Lime plaster, soft handmade brick, historic lime render and unheated masonry all rely on moisture moving out through the surface. A modern acrylic or alkyd film with low vapour permeability traps that moisture behind it.

The consequences are predictable: blistering and peeling within a few winters, salt crystallisation behind the film, and frost damage to the substrate it was supposed to protect. On external lime render the failure is faster and more visible.

The compatible options are the traditional ones and their mineral equivalents. Limewash and soft distemper on lime plaster internally, limewash or a mineral silicate system on lime render externally, and linseed oil paint on external joinery where the maintenance regime can support it. Where a client insists on a modern system, say plainly in the specification what the vapour resistance consequence is and record that the advice was given.

Lead is the other compatibility question, in the health and safety sense. Any pre 1970s scheme should be assumed to contain lead until tested, and its removal falls under the Control of Lead at Work Regulations 2002. Dry sanding and hot air stripping of lead paint on a listed building is both a health risk and a conservation risk, because it destroys the stratigraphy that the analysis has just recorded.

Reviewing a redecoration specification

A reviewer can work through a short list and catch most of what goes wrong.

1. Does the specification identify the substrate for every element, rather than describing "walls and woodwork"?
2. Does it state the existing coating type and whether lead is present or assumed?
3. Is the preparation method compatible with retaining historic layers, with stripping limited to areas where it is justified and recorded?
4. Are colours given as measured references with sheen and binder, not as trade names?
5. Are sample panels required on site, on the real substrate, with a named approval stage?
6. Does the scope of works match the consent, including any condition about retaining paint evidence?

The last is the one that reaches enforcement. Stripping historic paint from a listed interior can amount to alteration affecting the character of the building for the purposes of section 7 of the Planning (Listed Buildings and Conservation Areas) Act 1990, and consent granted for redecoration does not authorise removal of the evidence, which is why internal alterations that harm significance so often turn out to be works nobody thought needed consent. Reviewers who sign off specifications should look for that sentence and add it if it is missing.

This is general guidance, not legal advice; confirm the consent position for any works to a listed building with your local planning authority or a qualified heritage professional.

If you draft heritage documents with PlanScope, you can send the draft to a reviewer through the professional directory before submission. PlanScope provides AI-assisted draft reports for planning applications. Reports should be reviewed before submission. We do not guarantee planning approval. For specialist technical reports, qualified professional input may be required.