A laptop pulled from an upstairs bedroom after a house fire, scorched and then thoroughly soaked by what put it out. The platters were the least damaged part of it. The four-week delay was the expensive bit.
← All case files · £300 + VAT, flat
A family lost most of a house to a kitchen fire. Insurance, temporary accommodation and rebuilding took priority, and it was four weeks before anyone thought about the laptop that had been upstairs — which held every family photograph taken since 2009.
It arrived in a carrier bag: casing scorched and partly melted, screen destroyed, and the whole machine carrying the sharp smell of fire residue.
People assume the platters have cooked. They almost never have, and it is worth understanding why, because it changes what you should do in the first hour rather than the fourth week.
A hard disk stores data as magnetic domains on a thin film. Those domains survive until the material approaches its Curie temperature, which for the alloys used in drive media is several hundred degrees Celsius. The plastics, adhesives and screen around them fail far below that, so a machine can look catastrophically burnt while the sealed head-disk assembly inside never came close to losing anything.
The chamber itself helps. It is not airtight but it is filtered, sealed against particulate, and made of aluminium that conducts heat away from the platters faster than the surrounding case absorbs it.
What actually destroys these drives is the aftermath, and it operates on a completely different timescale to the fire.
No breach, no distortion of the casting. The platters had been protected from both smoke particulate and suppression water throughout.
Combustion produces acidic residues; add suppression water and you have an electrolyte sitting on exposed copper. Four weeks of that had eaten into traces and component legs, with visible green corrosion around several joints.
Every drive has a filtered vent equalising internal and external pressure. Fine residue had been drawn through it, though not in quantity sufficient to reach the flying surfaces.
No thermal distortion of the platters and no evidence of demagnetisation. Exactly as expected.
The four-week delay was the real problem, not the fire. Corrosion is a continuing electrochemical process, not an event: residue left on a board for a month does considerably more damage than the same residue removed within days, and the difference is not marginal.
Decontamination before any power. This order is not negotiable. Applying voltage to a corroded board completes circuits through conductive residue rather than through the traces, and the resulting current does in seconds what the corrosion had been doing over weeks.
The board came off and was cleaned in an ultrasonic bath with an appropriate solvent to lift the acidic residue, then rinsed and dried properly. Simply drying it — the instinct, and what a great many people do — concentrates exactly the material causing the damage rather than removing it.
Board repair. Several traces had corroded through and were bridged with fine wire; two components with badly attacked legs were replaced. On drives of this generation the board also carries adaptive parameters unique to the head stack in that particular drive, so the original board had to be repaired rather than swapped — a donor board without those adaptives will spin the drive and read nothing.
Chamber work. The drive was opened under filtered clean air to remove residue that had come through the breather and to replace the internal recirculating filter, which was loaded and would otherwise have shed contamination back onto the platters during the read.
One planned pass. With the board repaired and the chamber clean it was imaged read-only in a single carefully ordered sequence. A decontaminated drive is not a restored drive: corrosion damage to the head circuitry may only tolerate a limited amount of runtime, so the imaging plan assumed one opportunity and prioritised the photograph library’s regions accordingly.
About 88% of the drive, including the overwhelming majority of the photograph library. The losses were unreadable sectors scattered across the surface rather than concentrated in one region — consistent with intermittent read errors caused by degraded head circuitry rather than by damage to the media itself.
Photographs were verified by rendering and returned sorted by their embedded EXIF capture dates, which reconstructed a usable chronology spanning fifteen years even where folder structure had been lost.
Fire rarely destroys the platters. The sealed chamber protects them and the magnetic material tolerates far more heat than the machine around it. What does the damage is the aftermath — acidic smoke residue and suppression water, corroding continuously from the moment the fire is out until somebody removes them.
That makes these cases genuinely time-critical in a way most recovery is not. A clicking drive sitting in a drawer is in the same condition next month. A fire-damaged one is measurably worse every week.
So: do not dry it, do not power it, and do not leave it in a bag for a month while other things get sorted out. Seal it in a bag to stop it drying out further, and get it moving. Water and fire damage recovery is from £300 +VAT with a 50% deposit, after a free 48-hour diagnostic.
Drop the drive at our Oxford Street reception, or post it to us — it costs nothing to find out what happened. You get a written figure from the fixed bands before any work begins.