Recovery methodsTechnicalUpdated October 17, 2026 · Data Recovery Cases Editorial Team

Water, Fire and Contaminated-Drive Recovery

Environmental damage is a contamination problem before it’s ever a filesystem problem. Powering on, test-connecting or household-drying a wet, smoke-exposed or contaminated device can turn a condition a lab might still be able to handle into corrosion, electrical damage or head-media damage that wasn’t there when the device was first recovered — and what the device looks like from outside does not predict whether its data can be recovered.

In short

Do not power, charge or test-connect the device, even if it looks dry.
There is no universal wet-or-dry rule: follow the lab’s handling instructions and add no liquid.
What the device was exposed to matters, but composition and corrosion cannot be judged from a photo.
On HDDs, electronics and the sealed assembly are separate questions; an intact case does not prove a clean interior.
Professional stages may follow assessment, but not every device can be stabilized or fully imaged.

First decisions: what not to change

The first safe decision is mostly about what not to do. Do not apply power, charge a battery, or connect the device “just once” by USB or to another computer to see whether it still works. Each power-on of a contaminated board is a new event, and a device that appears to work briefly can still fail during the next attempt.

Do not put yourself at risk to retrieve a device for the sake of its data. If a building is flooded, fire-damaged or otherwise unsafe, recovery waits until the area is safe to enter; data is never a reason to go back into a hazardous area or to handle residues that may themselves be hazardous.

Do not try household or improvised treatments: no rice, heaters, hair dryers, ovens or freezers, no rinsing, no solvents or cleaning sprays, and no opening of the drive or its head-disk assembly. Equally, there is no universal rule that a wet device should be kept wet or should be dried. Published lab guidance itself varies — Ontrack’s own pages, for example, advise against drying a wet device before it is sent in — which is why the right step is to contact the lab you choose promptly and follow its specific handling and packaging instructions for that device and that exposure. Do not add water or any other liquid yourself.

What you can do without touching the device further is write down what is already known: which devices and media were affected, what they were exposed to and for roughly how long, and anything that has already been done to them since. Do not disassemble devices to find out which disks are inside; if that information isn’t already known, tell the lab it isn’t.

What you know, what it does not prove

Most early mistakes come from treating an observation as a diagnosis. The table separates what a person can usually observe from what it does not establish, and what that observation is not a reason to change.

Scroll horizontally to see the full table →

Common observations after water, fire or contamination exposure: what each does not prove, and what not to change because of it.
What you may know	What it does not prove	What not to change
The device looks dry and clean	That residues, moisture in connectors or hidden soot are gone, or that power is safe	Do not power it to “check”; do not wipe or clean it
The type of event (flood, burst pipe, fire)	What the liquid or residue actually contains, or whether corrosion has started	Do not rinse or neutralise it on the basis of a guess
A photo shows marks or discoloration on the board	The composition of the deposit, the extent of corrosion, or the state of the media	Do not scrape or brush it to get a better look
The casing is intact, or the HDD is a sealed helium design	That the head-disk assembly is clean, or that the device was waterproof or fireproof	Do not open it to confirm
The device was detected or spun up after the incident	That it is stable, or that further attempts are harmless	Do not repeat the test or start copying from it
The disk came from a RAID array or NAS	Which slot, member order or controller settings it had, unless those were already recorded	Do not move disks between enclosures or try them in another unit

Not all contamination behaves the same way

Relatively clean fresh water is a different problem from salt water, flood water, sewage, sugary or chemical liquids, or firefighting residue. Salt water can speed up corrosion, flood water is often dirtier and the exposure longer, sugary liquids leave sticky residue that is its own contamination problem, and fire incidents can combine heat, smoke, soot and suppression water or chemicals all at once. The type of contamination changes what a lab needs to do before anything else happens.

Those differences are a reason to report the exposure accurately, not to diagnose it. The name of the event says little about what a particular device actually met — floodwater, for instance, can carry very different contaminants — and neither the composition of a residue nor the degree of corrosion can be established from a photo. That assessment belongs to the lab, with the device in hand.

Why “it looks dry” does not mean it is safe to power

Liquid exposure can leave conductive, ionic, corrosive or otherwise harmful contamination on electronics and connectors. Drying may remove visible moisture while leaving residues behind. Electrochemical migration on a circuit board needs an electrolyte, such as adsorbed moisture, together with a voltage difference between conductors, and dissolved ionic contamination can promote it (He, Azarian and Pecht, CALCE). Applying power supplies the bias that was missing, so powering a device before it has been properly assessed or cleaned can create or accelerate leakage, shorts, corrosion and electrochemical migration. “It looks dry now” does not mean that process has stopped, and it does not mean the device is safe to power.

HDDs: external electronics and the head-disk assembly

This section applies to hard disk drives. The internal head-disk assembly (HDA) is a controlled recording environment separate from the external PCB, but HDD construction varies. Many conventional air-filled drives exchange air with their surroundings through a breather filter, while helium drives are hermetically sealed and filled with helium (Western Digital HelioSeal brochure). The two do not have identical environmental paths, so internal contamination risk depends on the drive and the exposure.

A sealed design is not a waterproofing or fireproofing rating, and an intact casing does not prove the HDA is clean: seals, labels, connectors and the board side of the drive can all be affected, and heat can deform parts that keep the interior closed. The media can sometimes be in better condition than a visibly corroded board suggests, and the reverse can also be true; neither can be judged from outside. Recovery has to treat external-electronics contamination and internal HDA contamination as separate questions with separate handling, rather than assuming that visible external damage tells you what happened inside.

SSDs, flash and memory cards: no head-disk stage

SSDs, USB flash drives and memory cards have no heads, platters or HDA, so there is no cleanroom head-and-platter stage in their recovery. The other risks still apply. Contamination on the board, connectors and components is an electronics problem in the same way, and powering a contaminated device carries the same kind of risk. Beyond the board, access can depend on the controller and its firmware, which maintain the mapping between logical data and physical flash, and on encryption that the device or the platform applies; damage to those parts can matter more than the condition of the flash chips themselves. SSD Recovery Fundamentals and Encryption Key Hierarchies & Recovery Metadata explain those dependencies.

Fire damage can be more than heat

A fire incident can combine several independent damage mechanisms at once — direct heat and deformation, smoke and soot deposits, and water or chemical suppressant used to put the fire out. Smoke damage is not always visible, and hidden soot can still cause problems once the device is powered (Ontrack, fire damage). Whether a residue contributes to leakage or corrosion depends on its composition, how it was deposited and the circuit conditions, none of which can be read from its appearance. Severe heat can damage electronics, lubrication, adhesives, heads, spindle/actuator tolerances, seals and enclosure geometry before anything can be concluded about the recorded information itself.

How badly a particular device was affected depends on the extent and duration of its exposure, and there is no external sign, temperature or time threshold that reliably predicts the outcome. Treat a fire-exposed device as both heat-damaged and possibly wet or contaminated until a lab has assessed it.

Keep original parts and storage records together

Keep the original PCB, any ROM or controller board, the enclosure and its bridge board, power adapters and cables with the device, even when they look damaged beyond use. On HDDs the original board can carry drive-specific ROM and adaptive data, so a replacement PCB is not a universal fix; that is covered in HDD PCB, ROM and Adaptive Data Recovery. For an external drive, access to the disk inside can depend on its enclosure’s bridge board or on encryption keys, so moving the disk to a different enclosure is not a neutral test (see Recovery Copies on source dependencies).

For RAID arrays and NAS units, keep any records you already have of slot positions, member order, controller and NAS settings, and do not move disks between bays, enclosures or units. Those records may be needed to rebuild the array from images later; Documenting RAID and NAS Systems Before Drive Removal explains what they cover. If labelling or removing the disks would mean handling contaminated equipment, leave that to the lab.

Training scenario: a NAS recovered from a flooded basement

Observed

A four-bay NAS stood in floodwater for some time. Once the area was safe it was taken out; it now looks dry. Someone removed one disk and tried it in a USB dock, where it was not detected.

Interpretation

The failed dock test does not show whether that disk’s board, its HDA or the dock was the problem, and it added one power event to a contaminated board. The original order of the removed disk may now be uncertain.

Safer next step

Stop testing, keep the remaining disks in the unit, note which bay the removed disk came from if that is known, and contact a lab for handling and packaging instructions for all disks and the NAS chassis.

Limit

Whether each disk can be cleaned, stabilized and imaged depends on its own condition after assessment; one disk in poor condition can limit what can be reconstructed from the array.

What professional handling may involve

After assessment, a lab may use some or all of the following, depending on the device and what it finds. None of these stages is automatic, and none is guaranteed to succeed.

Controlled cleaning and inspection. Electronics may need professional decontamination before any power is applied. On HDDs, where contamination has reached inside the HDA, cleaning and inspecting the platters and heads happens under the same particle-controlled conditions as any other internal HDD work (see Cleanroom Head Replacement and Donor Matching). That inspection is also diagnostic: a clean, undamaged surface calls for a very different plan from one with visible corrosion or scoring.

Parts work. Corroded or heat-damaged PCBs, connectors or head electronics can need repair or donor replacement, following the same compatibility and adaptive-data considerations as any other case; environmental damage adds contamination handling before that work, it does not replace it.

Stabilization and imaging. If the device can be made stable enough to read, it goes through the same controlled imaging as any other unstable source (see How to Image a Failing Drive Safely and Imaging Bad Sectors and Unstable HDDs). Not every damaged device can be stabilized, and a device that can be read may still yield only a partial image; recovery then works from what was read.

What to tell the lab

Accurate history helps the lab decide how a device must be handled before it is even unpacked. Useful information, where you already have it:

Which devices and media were affected, including removable cards and any disks you know were part of a RAID array or NAS.
What they were exposed to — fresh, salt or flood water, sewage, a drink or chemical, fire, smoke, soot or suppressant — and roughly for how long, as far as is known.
Whether the device was powered, charged, connected or tested after the incident, and with what result.
Anything done to it since: drying, wiping, opening, removing disks or swapping parts.
Any known storage configuration or encryption, such as RAID member order, NAS settings or whether the device was encrypted.
Whether the residues may themselves be hazardous, so the lab can tell you how to pack and ship safely.

Uncertain answers are fine when labelled as uncertain; a guess presented as fact can mislead the assessment. General advice to stop using a device, avoid do-it-yourself fixes, not open it and write down what happened is part of Seagate’s long-standing data-loss guidance as well (Seagate data-loss FAQ). For finding a lab, the providers directory is a reference list, not an endorsement.

The one thing not to do

Do not power on or test a device that has been wet, contaminated or through a fire, even once it looks completely dry, and don’t open, wipe, rinse, heat or freeze it yourself. Powering, opening or improvised cleaning can add new damage to a device a lab might still have been able to work with.

Technical references: Ontrack: water damage — one lab’s handling advice for wet devices · Ontrack: fire damage — do not power; hidden soot · Seagate Recovery Services: data-loss FAQ — stop using, do not open. These are companies’ own customer guidance, not industry standards; the Seagate FAQ is an older document. Sources on electrochemical migration and helium-sealed drives are cited inline.