In 2003, the International Maritime Organisation stated in 2003 that fewer than 1 in 20 alerts related to persons genuinely in distress:
1.4 False alerts in the IAMSAR Manual are defined as: Any alert received by the SAR system indicating an actual or potential distress
situation, when no such situation actually exists.
1.5 Due to an increasing problem it was decided to start collection of data on the causes for false alerts.
1.6 Statistics from (M)RCCs show that the percentage of false alerts are approximately 95-100% of the total alerts received, mainly caused
by lack of knowledge of the relevant conventions, codes and
regulations.
The European Radiocommunications Committee gives a breakdown of root causes:
Mishandling: 40%
Beacon Malfunction: 10%
Mounting Failure: 6%
Environmental Conditions: 4%
Unknown: 40%
It doesn't include figures for intentional, genuine alerts, but does list some cause descriptions that mostly indicate accidental, rather than frivolous, activation.
A report by the US FCC in 1995 states that around 10% of activations were genuine - this may indicate an increase in false alerts over that period (this one is interesting reading, with some notable examples in the text - start at page 10, labelled "18").
An undated presentation from US Coast Guard and NOAA says
- 96% of 406MHz EPIRB Alerts are false
- 85% Resolved by RCCs with registration and good detective work
Note that many of the above incidents measured above relate to automatically-activated beacons carried on boats, rather than PLBs, so it may be invalid to suggest that the PLB false-activation rate is the same as for EPIRBs in general. For example, the last document above says that
69% Of
False Alerts [are due to]
Bracket
Interface Failure
This clearly doesn't apply to PLBs.
A report that distinguishes PLBs from other EPIRBs is a January
2014 report by ICOA which says
Based on the data provided by Participants, Cospas-Sarsat calculates two false alert rates, identified for convenience as the “SAR false alert rate” and the “beacon false alert
rate”. The SAR false alert rate, which characterises the
impact of false alerts on SAR services, is the percentage of
false alerts plus undetermined alerts (no person in distress
fo und; no beacon found) over the total number of alerts
transmitted to SAR authorities. [...] In 2012, the false alert rate was 96%, i.e. about one
real alert in 25 alerts received.
In that year, the "beacon false alert rate" is listed at 0.4% for PLBs, compared to 0.9% for other EPIRBs and 4.9% for the old 121.5 MHz ELTs (i.e. one in 250 PLBs in the world transmitted a false alert; we can estimate that one in 5000 beacons was used in a real emergency in that year).