The Unique Challenge of Destroying Data on Solid-State Drives

Solid-state drives have largely replaced traditional hard disk drives in modern computing, offering faster performance, lower power consumption, and greater durability. But the same technology that makes SSDs better at storing data also makes them significantly harder to sanitise. The methods that work reliably on HDDs do not always work on SSDs, and organisations that fail to account for these differences risk leaving recoverable data on disposed equipment.

How SSDs Store Data Differently

Traditional HDDs store data on spinning magnetic platters. Data is written to specific physical locations, and overwriting those locations with new data effectively eliminates the original content. This straightforward relationship between logical addresses and physical storage locations is what makes HDD wiping relatively simple.

SSDs use NAND flash memory cells that store data as electrical charges. More importantly, SSDs use a component called the flash translation layer (FTL) to manage the mapping between the logical addresses that the operating system sees and the physical locations where data is actually stored. This abstraction layer means that when you tell an SSD to overwrite a specific logical address, the FTL may write the new data to a completely different physical location, leaving the original data intact in the old location.

This behaviour exists because of two key SSD characteristics: wear levelling and over-provisioning.

Wear Levelling and Its Impact

NAND flash cells have a limited number of write cycles before they wear out. To maximise the lifespan of the drive, the FTL distributes write operations across all available cells as evenly as possible. This process, called wear levelling, means that the SSD constantly moves data around internally without the operating system’s knowledge.

From a data destruction perspective, wear levelling means that simply overwriting all logical addresses does not guarantee that all physical cells have been overwritten. Old data may persist in cells that the FTL has retired from active use or that are being held in reserve as part of the wear levelling pool.

Over-Provisioning

SSDs contain more physical storage capacity than they advertise to the user. This extra capacity, typically between 7% and 28% of the total NAND, is reserved for wear levelling, bad block management, and performance optimisation. The user and operating system have no direct access to this over-provisioned space.

Data that was previously stored in user-accessible areas may have been moved to over-provisioned space as part of the FTL’s internal management. Standard overwrite operations cannot reach this space, meaning that remnant data may survive a conventional wipe.

Why Traditional Overwriting Falls Short

A traditional multi-pass overwrite, such as the DoD 5220.22-M three-pass method designed for HDDs, does not provide reliable sanitisation of SSDs. The overwrite operations are directed at logical addresses, but the FTL determines where data is physically written. There is no guarantee that the overwrite data will reach all physical locations where sensitive data may reside.

Research has consistently demonstrated that conventional overwriting of SSDs can leave recoverable data fragments in wear-levelled cells, over-provisioned space, and bad blocks that have been retired from active use. This does not mean that most of the data survives, but for organisations that need to ensure complete elimination, “most” is not good enough.

Effective SSD Sanitisation Methods

The NIST 800-88 guidelines address SSD sanitisation by recommending methods that work at the firmware level rather than relying on logical-level overwrites.

ATA Secure Erase and Enhanced Secure Erase: These are firmware-level commands built into the drive’s controller. When executed, the drive’s own firmware handles the erasure of all storage cells, including those in over-provisioned and wear-levelled areas. Enhanced Secure Erase is particularly effective as it applies a vendor-specific pattern to all cells. However, the reliability of these commands depends on the drive manufacturer’s implementation, and not all implementations have been independently verified.

NVMe Format and Sanitize commands: NVMe drives support standardised sanitise operations that provide better assurance than SATA-based secure erase commands. The NVMe Sanitize command supports block erase, crypto erase, and overwrite operations, with the drive’s controller managing the process across all physical storage.

Cryptographic erasure: If the SSD uses hardware-level encryption (self-encrypting drives), destroying the encryption key renders all data on the drive cryptographically inaccessible. This method is fast and effective, but it relies on the encryption having been properly implemented and the key being securely stored and managed.

Physical destruction: For the highest assurance, physical destruction of SSDs through shredding or disintegration eliminates any possibility of data recovery. The drive must be reduced to sufficiently small particles to ensure that individual NAND chips cannot be removed and read.

Verification Challenges

Verifying SSD sanitisation is more complex than verifying HDD sanitisation. After a software wipe, reading back the drive’s logical space will show that data has been overwritten. But this does not confirm the state of over-provisioned space, which cannot be read through normal interfaces.

Some forensic tools can access portions of an SSD’s raw NAND, but comprehensive verification requires specialist equipment and expertise. For most organisations, using a combination of firmware-level sanitise commands and post-wipe verification of the logical address space provides a reasonable level of assurance.

Choosing the Right Approach

Practical guidance: For SSDs containing standard commercial data, use firmware-level sanitise commands (ATA Secure Erase, NVMe Sanitize, or cryptographic erasure) through a certified wiping tool. For SSDs that contained highly sensitive or classified data, physical destruction to an appropriate particle size is recommended. Never rely solely on traditional overwrite methods designed for HDDs when sanitising SSDs. For guidance on building these distinctions into your processes, see our guide to building an IT asset disposal policy.

As SSDs become the dominant storage technology, organisations must update their data destruction practices to account for these fundamental differences. What worked for HDDs simply does not provide the same assurance for solid-state storage.

EWV provides NIST 800-88 certified data destruction for Victorian businesses, covering software-based erasure and physical destruction with full chain-of-custody documentation. Request a quote to discuss your requirements.