Transfer switch and mains

Count the outages, time the transfer, and know when the grid is the problem.

Every outage is a test of the whole chain: the switch, the generator start, the rectifier and the battery in between. The twin times each step of every outage and shows which sites ride through and which ones nearly don’t.

63 A 4-pole automatic transfer switch with mains-failure controller, 400 V, 50 Hz

Built and run to IEC 60947-6-1 (transfer switching equipment)

Start with one cluster of sites See how we prove results

Grid availability: 88.0%
Median transfer: 16 s
Outages a year: 320
Ride-through, new battery: 12.3 h

Anatomy and nameplate

1Mains voltage and frequency per phase. A phase below 80% counts as a mains failure.
2Switch position. Mains, generator or neither.
3Generator start demand. Sent 5 s after the mains fails.
4Transfer time. From mains fail to generator on load, every time.
5Battery voltage during transfer. The battery carries the radios until the generator takes over.

Transfer switch and mains nameplate
Switch
Rating	63 A, 4-pole, 400 V, 50 Hz
Interlock	Mechanical and electrical, never both sources closed
Mains failure	Any phase under 80% of nominal, or frequency 2 Hz off
Controller timings
Start delay	5 s
Settle before transfer	8 s
Retransfer delay	5 min of stable mains
Cooldown	3 min unloaded
Reference site
Outages a year	320 (85 under 3 minutes)
Hours without mains	1,054 h
Grid availability	88.0%
Median transfer time	16 s

Telecom site power chain

The grid or the generator feeds the site through the transfer switch. The rectifier makes −48 V DC for the radios and keeps the battery charged. The air conditioner keeps the battery cool.

Air
AC power
−48 V DC
Diesel
Has its own twin page

The physics

The equations behind every curve on this page.

Transfer time

t = start delay + crank + settle + switch

About 16 seconds on a healthy set.

Ride-through margin

battery autonomy ≫ start + transfer time

The battery has to hold the radios for longer than the generator takes, with room for a second start.

Grid availability

A = 1 − outage hours / total hours

88.0% at the reference site.

Outages and transfers

The twin counts every outage and times every transfer. A weak grid and a weak start battery both show up here first.

Outages per month and hours without mains

Reference site, 320 outages in the year

Outages
Hours without mains (right axis)

Grid availability: 88.0%
Outage hours: 1,054 h
Shorter than 3 minutes: 85

Each outage starts the generator and cycles the battery. Many short outages cost starts and wear even when the hours are low.

A = 1 − outage hours / 8,760

Transfer time, mains fail to generator on load

Every start in the year

Healthy start battery
Weak start battery

Median, healthy: 16 s
Median, weak: 23 s

A healthy set takes the load in under 20 seconds, every time. A weak start battery spreads the times out and adds second attempts long before the first failed start.

t = 5 s delay + crank + 8 s settle + switch

The battery in between

For the seconds between mains fail and generator on load, the battery carries the site alone. How far the voltage dips says how healthy it is.

Battery voltage through one transfer

50 A load, generator on load after 16 s

Healthy string
Aged string
String with a weak cell

Lowest, healthy: 49.6 V
Lowest, weak cell: 45.7 V

Every outage is a free battery test. The dip in the first seconds grows as the string ages, long before a full discharge would show it.

V(t) = N·(E_dis + (E_float − E_dis)·e^(−t/τ)) − I·R

Failure modes

What breaks, what it looks like, what the twin does about it.

Failure mode	The physics	Signals that show it	How the twin catches it
Slow or failed transfer	A slow start, a governor that will not settle or a contactor that sticks.	Transfer time longer than the site’s own history. Start demand with no generator on load.	Times every transfer and flags the ones outside the site’s normal spread.
Mains sensing not wired	The controller cannot see the mains, so outages and returns go unrecorded.	Generator runs with no recorded outage. Mains voltage that never moves.	Checks that the mains readings are present and sane before trusting any outage count.
Switch stuck on generator	Mains is back but the switch does not return, so the generator keeps burning fuel.	Mains healthy while the switch shows generator, past the retransfer delay.	Checks the switch goes back within the 5-minute delay and counts the hours when it does not.
Frequent short outages	A weak grid drops for seconds or minutes many times a day.	Many outages under a few minutes. A high start count for the run hours.	Counts short outages by site so the grid can be raised with the utility.

What the twin does not do

The twin does not operate the switch or change its timings. It shows each step of each outage, and your team changes the settings.

How we catch it early

A physics twin and a PINN, working on one transfer switch and mains.

The twin is a set of equations for this machine, tuned to its nameplate. Give it the flow, speed and conditions and it returns what every reading should be. It always gives the same answer for the same inputs.

A physics-informed neural network, or PINN, sits on top. It is trained on the plant’s own data and it is penalized whenever it breaks a physics rule. So it can only learn what the equations leave out: wear, fouling, friction. It cannot invent an answer that the machine could never produce.

Step 1
Measure
Flow, pressure, current, temperature, vibration. Whatever the asset already has.
Step 2
Physics twin
The equations of this machine, set to its nameplate. It says what the readings should be right now.
Step 3
Residual
Measured minus expected. Demand, weather and speed are already taken out.
Step 4
PINN
A small neural network that learns the cause of the gap, under the physics rules.
Step 5
Finding
A named cause, a date, and the kWh or hours at stake. An engineer reviews it.

loss = (measured − predicted)² + λ · (physics residual)²

The second term is the rulebook. It is zero when the network obeys the equations.

Why this beats black-box AI

A black-box model learns patterns from history. It needs examples of each failure, and it goes quiet when the pump runs at a point it has not seen. The twin starts from physics, so it works on day one. The PINN adds what physics cannot know about this particular machine.

That is also what makes the savings countable. The baseline is a calculation, not a guess, so the gap between baseline and meter is a number a customer can check.

What the twin checks on this asset

Every outage timed end to end
Mains sensing present and sane
Switch back to mains within the 5-minute delay
Battery dip at each transfer against the 16 s healthy transfer

	Fixed alarm	Black-box AI	Twin + PINN
Needs failure history to learn	No	Yes, lots	No
Says what is wrong	No	Rarely	Yes, by name
Works at a flow it has never seen	No	Often fails	Yes, the equations still hold
Puts a number on the saving	No	No	Yes, in kWh and dollars
Lead time on slow wear	Late	Varies	Weeks to months
Every number can be traced	Yes	No	Yes, to an equation

What the twin puts a number on

Energy and cost, from the equations.

Generator hours after mains came back

133 h a year

262 L with a 30-minute retransfer delay instead of 5

Sites under 10 minutes of ride-through

3 of 50

Modeled cluster: battery age, shelter temperature and load vary by site

Grid availability

88.0%

320 outages and 1,054 hours without mains in a year

What we do for you

We watch it. We catch it. We prove it.

Our reliability engineers run the twin on your transfer switch and mains, review every finding, and help your crew fix it. After the repair we compare the meter to the baseline and write the result in the ledger.

Start with one cluster of sites See the 90-day pilot

More assets

More from telecom towers, then the lead asset of each other industry.