False nuclear alarm in Ontario was due to human error, investigation finds
By Allison Jones
Officer immediately realized error, but uncertainties delayed correction
TORONTO (CP) — A false alarm about an incident at the Pickering Nuclear Generating Station last month was the result of human error, but a delay in sending an all clear was due to several systemic issues, a report found Thursday.
The alert was pushed to cellphones, radios and TVs across the province on the morning of Sunday, Jan. 12.
The duty officer at the Provincial Emergency Operations Centre is supposed to test both a live alert and a training system at shift changes, and on that day the officer thought they had logged out of the live system and into the training one when the alert was sent, according to a report from the chief of Emergency Management Ontario.
The officer immediately realized the error and asked supervisors how to fix it, but they were uncertain about whether or how to send a corrective alert to everyone who had seen the first, the report said.
“The findings revealed EMO procedural gaps, lack of training, lack of familiarity with the Alert Ready system and communication failures,” the report found. “These findings can provide context to the (duty officer) error and the length of time — 108 minutes — that elapsed between the alert issued in error and the second clarifying alert.”
Solicitor general Sylvia Jones acknowledged the systemic issues and said steps have already been taken to address them.
“As I did on Jan. 12, I unreservedly apologize for the alarm and anxiety caused to people across the province and I want to assure the public that everything possible is being done to prevent a similar event in the future,” she said in a statement.
“Emergency Management Ontario has already taken significant corrective action in key areas, including planning, procedures, operations, communications and staff training.”
The corrective steps taken include clearly labelling test messages in the alert system, requiring separate log-in credentials for the live and training systems, more training, and establishing a new procedure for an “end alert” message in case of future errors.