Was there some incident that triggered the interest in power and power stewardship at the Lab?

In 2019, we had PSPS events all over the Bay Area and elsewhere in California, which were planned outages by the utilities to prevent wildfires. These events served all of us because they made us realize that we need significant capital investments in the Lab’s power system.

Not only do we have an old and antiquated system that needs special care and maintenance, but thanks to AI, we will triple our power demand within the next few years.

What are the 3 Rs of power?

They are reliability, redundancy, and renewal. Reliability, we have to do a better job in maintaining the equipment; the redundancy aspect is that we need to improve the system architecture.

The laboratory’s power system was designed decades ago and no longer meets the demands of current mission needs. As a result, electrical faults—especially in medium- or high-voltage systems—can lead to prolonged outages. Maintenance activities are also challenging, as they often require widespread shutdowns and must be squeezed into short, inflexible time windows. This makes it difficult to service enough devices thoroughly, leading to insufficient maintenance and increased system risk.

It is important to emphasize that the laboratory’s power system is undergoing a modernization process through the implementation of a SCADA system—a substation automation system that will enable remote switching, automated and orderly system shutdown and restoration, predictive maintenance, integration of microgrids, and the development of a smart grid. This transformation will also empower the use of AI to support various aspects of system operations.

When many of the current systems were originally installed, safety regulations were much less stringent. At that time, it was acceptable to enter a common corridor and replace a failed cable directly. Today, that’s no longer allowed. Modern standards require fully independent systems to ensure that neither workers nor equipment are exposed to potential hazards such as arc flashes or short circuits.

So we need better architecture. Renewal involves replacing or consolidating outdated installations and ensuring that all new construction adheres to strict, modern specifications.

So we need better architecture. Renewal involves replacing or consolidating outdated installations and ensuring that all new construction adheres to strict, modern specifications.

How does this program work with LAMP?

LAMP is the first step in our Lab’s Electrical Infrastructure Master Plan. It covers the basics, such as replacing the antiquated main substation, Grizzly.

It would also improve the architecture of the system in some ways. But in addition to LAMP, some of the installations are 30 to 40 years old, and they’re running out of their life expectancy. Most existing medium-voltage (MV) equipment has exceeded its useful life. System consolidation (as opposed to direct replacement) is prioritized to streamline maintenance and enhance safety. New distribution loops will boost system reliability and redundancy, faster power restoration, and reduced operational disruptions during maintenance. Projects are identified and prioritized based on mission criticality, system integration, geographic location, and alignment with campus-wide initiatives.

All of this is critical if you want to have a Lab in 2045 that is functioning from a power system point of view.

What is the renewal portion of the program?

Let’s say you have one device that has reached the end of its life and another that is almost at the end of its life. Maybe replacing one for one is not the best solution. That’s what consolidation is about. It’s not technology that will help in making the decision to consolidate; it’s people. It’s where the engineers come in and shine. They look at the current devices and are innovative in looking for new solutions. 

Managing the Lab’s power system involves not just maintenance but also development. It has to be done in a circular and systematic way without consuming a lot of resources that are obviously needed elsewhere as well. It’s probably the optimal way to ensure that our system will work well and for a long time.