To answer your question, I reached out to Victor Vega, NXP Semiconductors‘ director of RFID solutions, who says that tag data would not be corrupted in such a scenario, though writing might not be completed. This, he notes, is not a common occurrence. Here’s an edited version of his response:
“Keeping the process simple, let’s just say that we wish to write to the lowest common denominator for length for a particular RFID chip. For most chips, that would be a ‘word’ (16 bits for Gen 2). And for the Gen 2 protocol, the write time is limited to 20 minutes per 16-bit word. Communications are established by the reader with the chip, and a memory bank is selected, along with the word destination. Let’s say the chip is within the RF field. The write process is initiated. If the tag is removed from the field after 10 minutes, there will not be enough time to complete the process of writing the word to memory. The memory, however, would not be corrupted—other words written previously would not be affected.
“This doesn’t happen in the real world that often. Readers are generally sophisticated enough to check that the memory was correctly written, and if not, they could retry writing that word until successful (or for a set number of attempts). Perhaps tags with longer data payloads may be more susceptible to incomplete writes if they need to be in the field for longer periods of time. But in general, all ICs use the same fundamental memory, so it’s basically the same process whether low-frequency (LF), high-frequency (HF) or ultrahigh-frequency (UHF).”
—Mark Roberti, Founder and Editor, RFID Journal
Can RFID Tags Store Arabic Alphanumeric Characters? »