A woman in Palmers Green rang me about her front bay last spring. Three cleaners had been at it before me. Each one told her the spotting would come off with the right product, took the money, and left the glass looking the way it looked when they arrived. By the time I got there the south-facing pane carried a bloom across the bottom third you could pick out from the pavement.
It had stopped coming off because it had stopped being a deposit.
That is the whole question on this kind of work, and nine jobs in ten nobody has explained to the customer that there are two separate answers hiding inside it. Mineral scale sitting on top of glass is a cleaning problem, solved in about eleven minutes a pane with the right acid and a flood rinse. Mineral scale that has sat on glass through four London summers is a glazing problem, and by then the cost of putting it right runs past the cost of a new sealed unit.
What Puts the Spots on London Glass in the First Place?
Chalk, mostly. The aquifer under this city gives us water at roughly 250 to 320 milligrams per litre expressed as calcium carbonate, and the boroughs on the Affinity supply out towards Ruislip and Northwood sit at the punishing end of that. Water lands on a pane, sits, evaporates, and leaves behind the calcium and magnesium it was carrying. Do that a few hundred times in the same spot and you have a ring with a hard edge and a raised centre.
The rain itself is rarely the problem. Rainwater is soft, and a pane that only ever gets rain will spot lightly and clean easily. The damage comes from water that has been somewhere else first.
I see four sources over and over. Irrigation overspray on podium schemes. Splashback off sills and window boards. Condensation running down aluminium and picking up whatever the frame has to give. And run-off from concrete, limestone and mortar above the opening, which is the one that ruins glass fastest, because that water arrives alkaline as well as hard, and alkalinity opens the silica up rather than settling on top of it.
Where the water has been before it lands
New-build podium gardens are the worst offenders I deal with. Half the schemes along the Nine Elms stretch irrigate off boreholes or harvested tanks rather than mains, which means untreated water with a higher mineral load and often a bit of iron in it, sprayed twice daily onto glazing at planter height by a head that was aimed at the shrubs on commissioning day and has been drifting ever since. Four seasons of that and the bottom 400mm of every ground-floor pane on the terrace side has gone opaque. The landscaping contractor’s sprinkler is doing more damage to that building’s glass than a decade of weather would.
How Do I Tell Deposit From Etch Without Guessing?
Look at the edges of the marks. Deposit reads as discrete rings and tails with a defined boundary – you are seeing dried droplets, and they have shape. Etch reads as haze with no edge at all, a cloudiness that fades out rather than stopping. Two panes in the same frame will often show you both, deposit high up and haze along the bottom where the water pooled longest.
Then feel it. Nitrile glove, light pressure, drag across the mark, and go slowly enough to register the change. Deposit catches. Etch is smooth, because the surface has been eaten into rather than built onto, and your fingertip glides over glass that no longer reflects properly.
Neither of those is proof. The proof is chemical, and it takes four minutes.
The corner test I run before I quote
I pick the least visible pane on the elevation – the return of a bay, the top light above a door, something facing a side return in Woolwich that nobody has looked at since 2011. Mask a 50mm square, apply a dilute phosphoric gel, leave it two minutes, agitate with a white nylon pad, flood off with pure water. Then dry it and look at the square in raking light against the glass around it.
If the square has gone clear, the whole elevation is a cleaning job and I can price it by the pane with confidence. If the square has lightened but not cleared, it is bonded scale and we are into cerium oxide and machine time. If the square is identical to its surroundings, the glass is etched and no product I own is going to change that. I photograph the test either way and I show the customer, because the conversation goes much better when they can see the boundary of the masking tape.
Any contractor who quotes hard water removal from a photograph is guessing, and nine panes in ten the guess flatters the job. I have seen a Croydon office block quoted at £4,200 for a restoration that a corner test would have shown to be impossible on about a third of the glass.
What Takes Scale Off Glass, and What Just Takes Money?
Fresh deposit, under about eighteen months, comes off with acid. Phosphoric or oxalic based, dwell time rather than pressure, white pad, and then more rinse water than you think you need – leave acid residue in a glazing bead and it will bloom back out in the first heavy rain and you will be back on site for free.
Bonded scale needs abrasion. Cerium oxide slurry on a rayon or felt pad, low speed, 800 to 1,200 rpm, and the glass kept flooded the entire time. Let the pad run dry for six seconds and the heat will put a lens in the pane that is more visible than the scale you were removing. On a 1.2 by 1.5 metre unit with heavy stage-two deposit I am budgeting somewhere between forty and ninety minutes of machine time, plus setup, plus protection to the frames and the reveal.
The sums nobody does out loud
Run those numbers at London labour rates and a single restored pane lands around £180 to £260. A replacement toughened double-glazed unit in an existing frame, same size, standard specification, is often under £200 fitted. I say this to customers knowing it costs me the job: on anything past stage two, restoration is a worse buy than glazing, and the trade sells it anyway because it books three days of work.
Etched glass – true corrosion, where the silica network has been opened up by alkaline attack – cannot be restored in any meaningful sense. You can polish it, which means removing glass until the pits are gone, which on a toughened pane with existing roller wave gives you a surface that distorts the view down the street. The customer notices that more than they noticed the haze.
Which London Glazing Must Never See Acid?
Self-cleaning glass. Pilkington Activ, Saint-Gobain Bioclean and their equivalents carry a titanium dioxide coating on the outer face, and acid or abrasive strips it in seconds. The patch you cleaned will then shed rain differently to the rest of the pane for the remaining life of the unit, which is a more obvious defect than the spotting you were hired to remove. Solar control coatings on face one, anti-reflective shopfront glass, and older applied films around the edge seal all fail the same way.
The maker’s mark and the O and M pack
On anything built since about 2008 there is a spacer bar stamp or an etched logo in a bottom corner giving you the glass specification. On a managed building, the operations and maintenance manual has the full glazing schedule and the facilities team can usually find it in an afternoon. Domestic stock is where this bites. On a job in Enfield with no paperwork and a window fitter who went out of business in 2014, you assume it is coated and you test in the least visible corner you can reach. That assumption has cost me a handful of jobs. It has never cost me a replacement unit.
Can Anything Stop It Coming Back?
Move the sprinkler heads. On podium schemes that single intervention removes most of the problem, and swapping pop-up spray for drip line at planter edges removes the rest. I put it in writing to the managing agent every time, and about half of them action it. The other half pay me to clean the same 400mm of glass twice a year for the life of the building, which suits my invoicing and offends me anyway.
The first eighteen months of a new build
The other window is the first eighteen months of a new build, when the concrete frame and the mortar joints are still leaching lime into every rainfall. Glass fitted during that period wants washing on a tight cycle, six to eight weeks, because the run-off has a pH high enough to begin etching within a season. Protection during the wet trades matters more than any restoration afterwards.
Sacrificial sealers – the silane and siloxane products sold as hydrophobic glass coatings – do work, in the sense that they buy you twelve to eighteen months on a sheltered domestic elevation. On a river-facing tower at Woolwich Reach they are gone by the second winter and you have paid a premium per square metre for the privilege. A shorter cleaning cycle with pure water beats every coating I have tested, and I have tested most of them.
Back to the Palmers Green bay. I ran the corner test, showed her the tape line, and told her the lower pane was finished and the upper two would clean. She had the one unit swapped by a glazier off Green Lanes for £190. The two I cleaned went back to clear in about half an hour, and I have been on that house every eight weeks since, which over three years has earned me considerably more than the restoration I did not sell her. The replacement pane has stayed clean because we took the downpipe splashback out at the same time.