It’s the Smart Grid, Stupid

What do a Swiss company, the 1992 election, and a Berkeley professor have in common?

Nothing really. I’ve just been found wanting for a catchy title.

Pacific Gas & Electric is in talks with Switzerland’s Landis + Gyr and General Electric to roll out a $1.7 billion advanced metering infrastructure program over a three to four year period. Though the three have yet to sign a contract, a “person with knowledge of the negotiations” expects ink to meet paper sometime in the next two weeks. Landis + Gyr and GE are expected to share in the AMI deal because, as a PG&E smart energy director Andrew Tang said, “it makes a lot of sense to have more than one supplier…(especially when) we’re going to be deploying something on the order of 10,000 to 12,000 meters per day.” The two companies will divy up the reported $450 million metering component of the contract. All in all, not a huge pay day for either firm.

And that other $1.25 billion? Some of it will go to Silver Spring Networks, which, incidentally, signed a deal this week with PG&E to supply five million of the company’s customers with network infrastructure, devices, and operating software. PG&E provides electricity to about 15 million people, so this gives us a good idea of the breakdown between the other two firms. Silver Spring Networks says the new infrastructure will come online by 2012.

One the clearest solutions the building up of a smart grid is likely to effect is an increase in the relative availability of price signals within electricity markets. Large industrial customers in most markets have access to real time prices, often because those customers account for a significant, though not disproportionate, amount of load. Companies like Comverge and EnerNOC have taken advantage of this to an extent with their industrial demand response programs. For retail customers, however, these signals are rarely transmitted. Our bills are adjusted twice - maybe three times - a year to reflect the resuts of ratemaking negotiations between retail suppliers and utility commissions. I’ve heard those hearings are worse than prison and, differently than prison, you’re not likely to make any new friends.

As such, retail consumers can be described as relatively price inelastic. They’ll pay for electricity regardless of the cost - and not because it’s priced in Travis dollars. It’s because they don’t know what the true cost of electricity is. Now, imagine you had a little device in your house to tell you electricity just moved shot past $0.50/kWh and maybe it would be a good time to turn off two or three of your eight TVs. Good, right?

Maybe, says Berkeley professor Severin Borenstein. You may remember Prof. Borenstein from such hits as “The Market Value and Cost of Solar Photovoltaic Electricity Production” and Harry Potter. Borenstein wrote a paper in 2005 describing the efficiency gains derived from real time pricing. Price responsive customers tend to exhibit higher elasticities of demand, and will respond more readily to available price signals. However, as users are added to the system, efficiency gains from demand reduction begin to move out at the margin. Because most industrial customers - those that have shareholders that require them to be price responsive - are already on some form of real time pricing, we would find more gains from improving service to them then through extending service to retail customers.

According to Borenstein, this is because retail customers are lazy. Or, at least, not as responsive to price movements. For the utility that has to pay to roll out these devices ($1.7 billion doesn’t grow on trees), betting that its customers will turn down their power usage may prove a risky one. The investment only really pays off if demand and usage drop.