TowerCo · Case study

TowerCo battery, SLA & genset.

Three scenarios from a regional tower portfolio — fuel drift, a carrier SLA dispute, and a battery that would have failed at hour four of a nine-hour storm.

A TowerCo running 600+ unstaffed macro sites across the central US. Carrier-tenant SLAs, generator runtime during storm season, and battery strings that vendor BMS rounds to "healthy." Three incidents ObservOne handles on one console.

Read the scenarios
9 daysgenset lead time — fuel flag to landfall
47 → 4minutes of disputed carrier downtime
9h 12mstorm backup held on a swapped string
Operator scenarios

How this plays out in the field.

01

A fuel run, not an outage call.

OKC-TWR-14 · genset fuel rail · sustained −2.8 PSI · 6h
01
OKC-TWR-14 · genset fuel rail · sustained −2.8 PSI · 6h

Pressure rail drifts while vendor says OK.

Vendor RMS shows fuel “OK” — the float is sticking high. Fuel pressure at the rail has been running 2.8 PSI low for six hours across three consecutive 30-minute samples. A storm system is forecast in ten days. The genset will be asked to run.

02
correlation engine · Modbus vs tank float · disagreement flagged

Console correlates against runtime profile.

ObservOne cross-checks the Modbus pressure reading against the controller’s self-reported “OK” status. Three consecutive windows below threshold, each contradicting the vendor. Diagnosis: probable fuel-filter restriction. Intervention pillar pulls the already-scheduled O&M loop and flags the fuel-system check.

03
Intervention · P3 · fold into Tuesday maintenance loop

No truck added. Storm passes clean.

On-call sees one line: “OKC-TWR-14, fuel rail drift, no transfer risk yet, fold into Tuesday’s loop.” Fuel filter replaced during the scheduled run. Nine days later the genset holds four carrier tenants through eleven straight hours of grid outage. No SLA timer ever starts.

9 days
of lead time
From pressure-rail flag to storm event — enough for a scheduled maintenance fold, no added truck.
4
carrier tenants protected
No utility outage reached the BTS aggregator. No carrier SLA timer started.
FCC Part 17Carrier SLA
02

The screenshot lost. The sealed stream won.

MIA-TWR-08 · MNO-C dispute filed · −47 min uptime claim · Q2
01
MIA-TWR-08 · MNO-C dispute filed · −47 min uptime claim · Q2

Carrier claims 47 minutes down.

MNO-C files a Q2 SLA dispute worth ~$38K claiming 47 minutes of carrier-affecting downtime during a Miami thunderstorm. The TowerCo vendor RMS shows the site as “available.” The carrier has screenshots from their own NOC showing outage. Both sides are right — they’re measuring different things.

02
audit log · Ed25519 sealed · per-tenant impact

Sealed stream shows 4 minutes 9 seconds.

ObservOne’s hash-chained log captured every state transition: utility loss, genset transfer, BTS power-good, sector RF re-up — each Ed25519-signed at write time. Applied to MNO-C’s specific sector: 4 minutes 9 seconds of RF unavailability. The remaining 43 minutes were MNO core-network reconvergence on their side of the demarc.

03
dispute packet · delivered · MNO-C withdraws in 72h

Dispute resolved on evidence, not negotiation.

Compliance pillar drafts the response: sealed timeline, per-tenant impact breakdown, raw event stream attached. MNO-C withdraws 43 of 47 claimed minutes within 72 hours — ~$35K retained. Their NOC now knows the operator’s evidence is dispute-grade. Next quarter’s dispute volume drops measurably.

47 → 4
minutes of disputed downtime
43 minutes withdrawn in 72 hours. Carrier core reconvergence was on their side of the demarc.
~$35K
SLA credit retained
Sealed event stream settled what a screenshot could not.
Carrier SLAEd25519 audit log
03

The site that should have died at hour four.

TPA-TWR-21 · battery string B · projected runtime 4.2h · class A1 storm in 5d
01
TPA-TWR-21 · battery string B · projected runtime 4.2h · A1 storm in 5d

BMS says healthy. Model says 4.2 hours.

Vendor BMS reports “healthy.” But the per-string aging model shows three consecutive discharge cycles creeping down — 7.4h, 6.8h, now projecting 4.2h under the next storm’s load profile. The site’s SLA commitment assumes 8 hours. A Cat-1 system is forecast to make landfall in five days.

02
Power & Battery pillar · pre-storm priority elevated

Swap dispatched before the storm window.

Power & Battery pillar flags the site as pre-storm priority. Intervention pillar pulls the nearest tech, the regional depot’s spare bank, and the carrier SLA exposure if the swap is missed. Runbook attached. Operator approves with one click. Four days before landfall.

03
swap complete · storm event · 9h 12m backup · all RF nominal

Carrier never knows the grid was down.

Grid drops 41 hours after the swap and stays down for 9 hours 12 minutes. The site holds the full BTS load through the entire window. Three adjacent portfolio sites also flagged early — two swapped, one runbook’d. No carrier tenant on the affected network loses RF.

4.2h
projected runtime pre-swap
Against a 9-hour storm event and an 8-hour SLA commitment to carrier tenants.
9h 12m
storm backup duration
Site held the full BTS load through the entire window after the swap.
Carrier SLAFCC Part 17
Further reading
TowerCo operations5 min readMay 2026

Why 60% of tower truck rolls clear nothing — and what to do about it

Single-reading alarms, no cross-correlation, vendor RMS limitations. The three structural failures behind no-fault-found dispatches — and how proper alarm correlation fixes it.

Read →
SLA & compliance evidence5 min readMay 2026

Carrier SLA disputes: why evidence wins and screenshots lose

Screenshots from vendor RMS lose SLA disputes. Hash-chained, Ed25519-signed event streams win them. Why the gap between testimony and evidence is wider than most operators realize.

Read →
Hands on

Reproduce this scenario
in our sandbox.

30 minutes with a solutions engineer. We'll preload a tenant with anonymised sites matching your topology. NDA-friendly.

Book a demo All use cases