The Siren Records Came Back. The Answers Stayed Silent.
Follow-up to “The Tornado Siren Shuffle: Jennings County Edition”
Back in June, Granny wrote about Jennings County’s tornado sirens staying silent during two tornado warnings on June 17, 2026.
At the commissioners meeting, officials said the failure was “obviously an electrical failure.” They floated a lightning strike or an electrical surge. The public was told that “three pieces of viable equipment don’t go down at the same time typically unless there’s an electrical surge.”
We were also reminded that sirens are not the primary warning system, do not cover the entire county, and should not be relied upon indoors. Those things may be true, but they did not answer the question folks were asking:
Why did the emergency warning system Jennings County owns, maintains, tests and pays for fail when it was actually needed?
Around the same time, a resident requested the records via APRA.
They finally arrived, they shared it.
The records do not prove the county’s public explanation. They mostly prove the county had a theory, an aging system and a paperwork drawer with some important pages missing.
The Official Diagnosis Came From “No Electrician”
The centerpiece of the response is a file named **County IT.pdf**.
It is one page. It has no author’s name, no signature and no date written on it. The PDF metadata shows it was scanned on June 24, one week after the siren failure.
The writer says they accompanied the 911 director to the radio room on “Monday morning.” By then, the radio transmitter had already been checked by “the radio repair people” and removed as inoperative.
The writer observed that two APC battery-backup units would not power on. The outlets worked. The grounding wires appeared tight. A backup transmitter was plugged into a third working battery backup.
Then comes the county’s technical conclusion:
“I am no electrician but the most likely cause to in my opinion is the Radio antenna or the tower the antenna is on was possibly struck by lightning.”
There it is.
The county publicly presented an electrical surge as the obvious explanation. The records supplied to support that explanation contain an unsigned memo from an unidentified person who openly says, “I am no electrician,” followed by “most likely,” “in my opinion” and “possibly.”
That is not a professional diagnosis. That is a theory wearing an IT badge.
Lightning may have caused it. A surge may have traveled down the antenna line. Nobody is saying that is impossible.
But the response contains no lightning report, no electrical inspection, no surge analysis, no equipment diagnostics, no photographs, no qualified electrician’s findings and no final report from the radio technician.
The county gave the public a conclusion. The records give us a guess.
The Repair Technician Apparently Vanished From the Paper Trail
At 12:30 p.m. on June 18, the day after the warnings, 911 Director Britt Burgmeier emailed ERS Wireless:
“We had 2 tornado warnings last night, but our sirens would not sound.”
Burgmeier said troubleshooting that morning showed the repeater in the back room had no power. Different cords and outlets had been tried without success.
ERS scheduled a technician for 10:00 the next morning, June 19.
And that is where the supplied vendor records stop.
There is no June 19 technician report. No service ticket. No written diagnosis. No invoice. No list of damaged equipment. No explanation of what testing was performed. No documentation showing when the system was restored or whether every siren was successfully activated afterward.
The response tells us a repairman was coming.
It does not tell us what the repairman found.
That missing report ought to be the most important record in the entire packet. Instead, the county supplied an email appointment and left the actual diagnosis standing outside in the rain.
The “Backup” Was Not Ready to Back Anything Up
The IT memo says a backup transmitter was available and was plugged into a third working battery backup after the original transmitter and two APC units were found inoperative.
That raises another question.
Was the backup transmitter already operating as an automatic or immediately available redundant system on June 17, or was it sitting there waiting for someone to discover the failure, move equipment around and plug it in afterward?
A spare part in a room is useful. It is not the same thing as functional redundancy.
If the entire countywide system depends upon one transmitter, one antenna path and battery backups sitting beside each other, then a single surge can apparently knock out the primary equipment and its backups together.
That is not three independent protections. That is three eggs riding in the same electrical basket.
The Written Failure Procedure Is One Sentence
The county provided its Tornado Siren Operations policy. It became effective in 2011 and was most recently reviewed on April 4, 2024.
The activation procedure is simple. Dispatch clicks “TOR SIREN (ALL SIRENS)” and then accepts and sends the command.
The failure procedure is even simpler:
“If the tornado sirens fail to activate, notify the director immediately.”
That is the entire written backup procedure supplied.
There are no instructions for switching to the backup transmitter. No alternate activation procedure. No manual activation instructions for individual sirens. No list of township or fire-department contacts. No checklist for verifying which sirens activated. No written public-notification fallback specific to a systemwide failure.
Click the button. If nothing happens, call the director.
That is not much of a contingency plan for a countywide emergency warning system.
Where Are the Weekly Test Records?
At the commissioners meeting, officials said quiet tests were routinely conducted on Mondays. They said the test sends a signal, performs a small rotation and audio burst, and “gives us a report back.”
The last quiet test before the storm was reportedly two weeks earlier. The next one was missed because Burgmeier was on vacation.
If the system gives a report back, those reports should be records.
None were included.
The packet contains no weekly test log, no activation history, no console report, no failed-test notice and no record showing the condition of the system between the last successful test and the June 17 failure.
There is also no dispatch log showing the attempted activations during the two warnings.
Maybe those records exist elsewhere. They were not in this response.
That matters because the county’s public defense relied heavily on regular testing. If testing is the proof that the system was healthy before the storm, the testing records should not be the invisible man at the records party.
The Latest Siren Inventory Is Eight Years Old
The only site inventory provided is labeled:
“SIREN PROJECT-JENNINGS-12 current to 4/1/2018.”
It lists 12 siren locations.
That was eight years before the 2026 failure.
The response contains no current inventory identifying the sirens operating in June 2026, their manufacturers, models, serial numbers, installation dates, battery ages, inspection dates or current condition.
For an emergency system spread across the county, “current as of 2018” is doing heroic work for the word current.
The 2023 Records Showed Trouble Across the System
Four ERS Wireless invoices from 2023 total $11,589.50.
Those invoices show that at least seven of the 12 listed locations had a documented mechanical, antenna or battery concern:
The Vernon Gym siren had a failed horn motor and had to be removed for factory repair.
Spencer Township had no rotation, a bad antenna and all four batteries failed their load test.
Geneva had a bad antenna and all four batteries failed their load test.
Country Squire Lakes had no rotation and swollen batteries.
Campbell Township had a bad antenna.
Marion Township had two failed batteries, with replacement of all four recommended.
The EMA Training Facility batteries passed the load test but dated to 2017 and were recommended for replacement because of their age.
Temporary magnetic mount antennas were installed at Spencer and Geneva. The records supplied do not show when permanent antennas replaced them.
The packet also does not contain later invoices or work orders establishing that all the recommended batteries, antennas and non-rotating sirens were repaired.
That does not prove the problems remained unresolved through 2026. It proves the response did not document their resolution.
There is a difference, and the county ought to be able to close it with records.
Even the Billing Raises Questions
The Vernon horn removal invoice says the work “took 3 guys and 1 hour,” plus one hour for a technician to prepare his report.
The invoice billed four hours each for three installers, another technician hour and three separate $78 service-call charges, totaling $2,149.
Travel and mobilization may explain some of that billing. The invoice does not explain it.
The original repair estimate was $2,074. The removal and later reinstallation invoices together totaled $4,631.
That may all be legitimate. It is still worth asking why the final cost reached more than twice the original estimate and why three service call fees appeared on one removal invoice.
Public money should not require interpretive dance.
The Public Explanation Has Not Been Proven
At the meeting, the public was told the incident could not have been prevented or predicted and that everybody did what they were supposed to do.
The records do not support such a broad declaration.
They show the sirens failed during two warnings. They show a transmitter without power. They show two dead battery backups. They show a spare transmitter being connected afterward. They show a system with documented battery, antenna, rotation and motor problems in 2023.
They do not show the final technical cause.
They do not show the weekly testing history.
They do not show the June 19 technician’s findings.
They do not show when the full system returned to service.
They do not show a current siren inventory.
They do not show that the earlier site problems were permanently corrected.
And they do not show a serious written contingency plan beyond notifying the director after the button fails.
The county may eventually produce records that answer every one of those questions. This packet does not.
Granny’s Side-Eye
The first time around, Granny said “electrical failure” was not an explanation.
Now the records have arrived, and the county’s “obvious electrical surge” still rests primarily on an unsigned memo from somebody who begins the conclusion with:  “I am no electrician.”
Meanwhile, the actual radio technician’s report is missing, the test reports are missing, the dispatch activation records are missing and the latest siren inventory is old enough to be in third grade.
Jennings County’s sirens may not be the primary warning system.
They are still a publicly funded emergency warning system.
The public did not demand magic. It demanded maintenance, redundancy, records and an honest explanation when the system failed.
So far, the excuse machine has produced more noise than the sirens did.
Sherry Shepherd
 ·
Do people actually depend on these sirens to alert them of bad weather? At night when we're asleep and with the air conditioner running,we don't hear it. During the day we've heard Hayden's siren. Don't most people get the alert on their phone?
Reply
Send message
Hide