73. Defining Flexibility: How Flex MOSAIC™ Creates a Shared Language for the Grid

So why don't we sort of set the stage for the rest of the episode and Anuja, for listeners who may be new to DC Flex and Flex Mosaic, can you explain what Flex Mosaic is and what problem it was designed to solve for?

Absolutely.

So DCFlex, the challenge we were going after with DCFlex was, can we unlock existing grid capacity using flexibility as a characteristic of large loads? And Flex Mosaic is really what defines what that flexibility is.

So fundamentally, Flex Mosaic is a common language classification that gives people performance-based requirements on how to bring flexibility to bear.

And the reason this was important was because during the last year, as we were engaging with our members in DC Flex, We kept hearing this conversation that was a circular conversation where a data center would come to the table and say, I need to be connected.

And a utility would say, well, to be connected in the timeframe you need, we would be able to make it happen if you were flexible.

Then the data center goes, what do you mean by flexibility? What do I need to bring? And utility goes, tell me what you can bring to bear and I might be able to tell you if that's a good donor.

So in order to break this cycle, we need a clear characteristics of what flexibility is.

And that's what FlexMosaic brings to bear.

It's a framework that's really quite simple.

It essentially categorizes grid needs by classes with very clear performance parameters like notification time, activation periods, annual utilization.

So when a data center gets the directional guidance from a utility that says, if you want to be connected at this location, in under three year window, you need to bring to bear class A flexibility as part of your design.

What that means to the data center is you need to be able to take a notification for a grid event.

That's a day ahead notification.

You need to be able to be flexible and reduce your load for up to five hours.

And if it's class A, you would be seeing that kind of grid events about 1% of the year.

So less than 90 hours in the year is when you would see this.

It's as simple as that.

That's what the classification brings.

But there are five classes that addresses sort of the vast majority of grid needs that's driving the interconnection timing.