The sub-millisecond surge that can sink your AI investment: Why power quality is now a boardroom issue

A maintenance manager had a puzzle he could not solve. A PLC on his line had failed, been replaced, and failed again. By the third failure, his team had checked everything: no overload, no short circuit, no wiring fault.

“We have replaced this PLC twice this year, and now it has failed again,” he told engineers at Schneider Electric.

The culprit, found only after weeks of investigation, was a transient voltage surge lasting less than a millisecond. It was far too brief for ordinary meters to register, yet powerful enough to destroy hardware again and again.

For Amit Sharma, Vice President – Power Products and Digital Energy, Greater India Zone at Schneider Electric, the story captures why power quality has become one of the most underestimated risks in enterprise technology, and why CIOs, CFOs and boards can no longer treat it as someone else’s problem.

The context: AI is raising the cost of every disturbance
India is building at speed: AI infrastructure, hyperscale data centres, semiconductor ecosystems, advanced manufacturing, EV charging networks and large-scale renewables. Each depends on electricity that is not just available, but clean and stable.

That is a different demand from the one the grid was designed around. Sharma argues that digital transformation has changed what a disturbance costs.

“A data centre, a semiconductor process or an automated production line has very little tolerance for disturbance, and an event lasting a fraction of a second can bring operations to a halt,” he says. “Power quality is no longer an electrical issue. It is a business continuity issue.”

He sees national stakes as well. “As India builds advanced manufacturing, semiconductor ecosystems and AI infrastructure, power quality will increasingly become a competitive advantage rather than just a technical requirement.”

The irony is that the technologies driving digital growth are also adding to the problem. “Drives, UPS systems, EV chargers, data centres and solar inverters are introducing new disturbances into the electrical system,” Sharma notes.

Compliance is a financial metric, not a paperwork exercise

With kVAh billing and IEEE 519 harmonics standards gaining ground across India, Sharma says many enterprises still misread what is at stake.

“The most common misconception is that compliance is only a regulatory obligation,” he says. “Power factor and harmonics are no longer engineering metrics. They are financial metrics. kVAh billing places a cost on power factor, while IEEE 519 sets discipline around harmonics.”

Another common error is seeing IEEE 519 purely as a way to dodge penalties. In Sharma’s view, it is “a standard for electrical discipline, and compliance is about protecting the health of electrical infrastructure.” And the effects do not stop at the factory gate: a harmonics problem at one facility can become a reliability challenge for the wider network, which is why utilities and policy makers pay close attention.

A further blind spot is fixating on power factor correction while ignoring harmonics from modern electronic loads. The bill for that oversight arrives in disguise: higher energy costs, nuisance tripping, overheating, and shorter life for transformers and capacitors.

“Power quality problems are often invisible until they become expensive,” Sharma warns.

The quiet ROI killer
Unlike an outage, poor power quality rarely announces itself. “Poor power quality quietly erodes profitability long before it causes failure,” Sharma says. Energy losses, equipment wear, rising maintenance costs and falling efficiency compound over time. It is, he says, “one of the few electrical challenges that affects both the top line, through production continuity, and the bottom line, through energy efficiency and asset reliability.”

That is where the PLC story lands. Events that brief cannot be seen by standard instruments. “You cannot control what you cannot measure,” Sharma says, which is why he champions Class A power quality metering. It captures such events so the root cause is found early, “and not after repeated failures.”

Warning signs technology leaders should not ignore

Sharma points to a short list of early signals:

Frequent breaker or capacitor failures
Repeated equipment tripping
UPS alarms
Overheating assets
Unexplained increases in energy consumption

“The electrical system is speaking all the time,” he says. “Monitoring helps us listen.”

Renewables, EVs and the next wave of grid stress
India’s progress toward One Nation, One Grid, One Frequency has created a unified electricity ecosystem. But Sharma cautions that frequency synchronisation is only one piece of reliability.

As renewables, EV charging, data centres and distributed energy resources scale, they bring “voltage fluctuations, harmonics and dynamic load changes that networks were not originally designed to handle.”

His framing is useful for anyone planning a capital programme: “Power quality acts as the bridge between grid availability and grid usability.” Power must be delivered “in a form that sensitive industrial processes, AI infrastructure, healthcare facilities and advanced manufacturing can reliably use.”

The advice for businesses is to act before deployment. Assess how ready your electrical infrastructure is, and put continuous monitoring in place before new assets go live. Problems then surface early, “allowing new energy technologies to be adopted without compromising reliability or efficiency.”

A lifecycle approach: Design, Monitor, Analyze, Act, Optimize
Sharma insists power quality should be “designed in from Day 1, and not corrected after problems appear.”

Schneider Electric treats it as a lifecycle:

Design power quality into the electrical system from the start.
Monitor with Class A meters.
Analyze with software such as Schneider’s Power Monitoring Expert (PME).
Act through a PQ audit by certified auditors, or through analysis of PME data.
Optimize with corrective measures such as power quality equipment.

The goal, he says, is to be “a consultative, end-to-end power quality management partner, from the first design decision to the last corrective action.”

Five years out: from firefighting to a strategic edge
Looking ahead, Sharma expects intelligence to replace manual intervention. “With always-on data and AI-based insights, organizations will move from reacting to failures to condition-based maintenance,” he says, improving asset reliability, cutting downtime and optimising maintenance spend.

His mantra is Measure, Analyze, Improve. “Visibility is the first step toward reliability,” he says. “What gets measured gets improved. What gets ignored becomes a risk.”

There is a sustainability payoff too. Better power quality is “one of the fastest ways to improve electrical efficiency without adding new generation capacity.”

Sharma’s closing message is a challenge to the way enterprises think about energy: “India has largely solved the challenge of power availability. The next challenge is power quality.”

For technology leaders, the practical questions are plain: Do you have Class A-level visibility into the power feeding your most critical systems? Are harmonics and power factor tracked as financial metrics? Was power quality part of the design brief for your last data centre, plant or EV rollout?

Sharma’s view is that the answers will shape competitiveness. “The future competitiveness of our industries, data centres, EV infrastructure and digital economy will increasingly depend on the quality of every unit of electricity delivered.”