The distribution power grid -- the part of the grid that carries medium- and low-voltage electricity from substations to end users -- is mostly “dark” nowadays, with little in the way of sensors or data for utilities to use to map out just what problems those parts of the grid are facing, let alone how to fix them.

Bringing light to that darkened distribution grid is the goal of the sensors, communications and control systems that make up distribution management and automation systems. DA is a big and growing business -- GTM Research predicts the U.S. market alone will grow from an annual $1.75 billion in 2012 to peak at just more than $3 billion in 2015.  

All of that gear is designed to sit out on the grid for decades, hopefully returning its value in improved grid efficiencies and new capabilities over time. But what if a utility could move its networked sensors from one section of the grid to another, testing each out, discovering its inefficiencies and fixing them, and then move on?

That’s the idea behind Awesense Wireless, a Vancouver, Canada-based startup founded in 2009 that’s landed some significant utilities as customers, including hometown Fortis BC and Brazil’s ELO Sistemas Electronicos late last year.

Awesense, which announced last month that it raised an undisclosed round of funding, has mainly deployed as a turnkey service to date, selling utilities its equipment and training workers to go out and deploy and manage the system. On Monday, it launched a service-based version of the business, which will allow utilities to pay for the deployment via a revenue-sharing agreement -- and, potentially, allow the startup to expand to customers that might be leery of investing in all the gear upfront.

The company already has one U.S. utility with approximately 34,000 meters and 1,800 miles of lines using senseNET under just such a service contract, CEO Mischa Steiner-Jovic said in a phone interview. It’s also working with utilities in Malaysia, Turkey, Bulgaria, and the Czech Republic, which plan to begin trials in the first quarter of this year.

“The distribution portion of the grid is the most inefficient portion of the grid,” Steiner-Jovic said. Today, utilities use antiquated methodology to find these losses, based on math that tries to get around the lack of data, he said. “We’ve automated the entire process for them, and added a lot of technology,” he said. “In one or two days, they can identify where the loss is occurring, as opposed to weeks and months of analysis.”

Awesense’s senseNET system combines line sensor devices, which can be dropped onto distribution cables by utility workers using so-called “hot-stick” methods -- no climbing poles and rewiring grid circuitry required. A typical project might include several hundred such sensors, put out over the course of a few days and connected via the company’s own 900-megahertz band mesh networking.

Once they’re up there, the devices self-network and draw back tons of data on power quality, which feeds into the senseNET software suite that helps identify key inefficiencies. Those can include technical losses, such as worn-out transformers or sections of current experiencing phase problems -- but it can also include so-called “non-technical” losses, which is a fancy term for power theft.

Awesense first tested its system with utility Fortis BC, and while it hasn’t officially named BC Hydro as a customer, it did win a $40,000 sustainability prize from British Columbia’s biggest utility in 2011. The company is also working with unnamed advanced metering infrastructure (AMI) vendors to integrate its data into their systems, Steiner-Jovic said, though he didn’t name the partners involved. Both Fortis BC and BC Hydro are using Cisco and Itron for their smart meter deployments.

Many of the utilities working with Awesense are rolling the costs into their grid operations and management budgets, he added -- a move that allows utilities to avoid having to make a rate case for capital investments. That’s a testament to the fact that deploying sensors to gather data, then moving them on to the next section of the grid, is a lot cheaper than installing a lot of permanent sensors on all those portions of the grid. Just how much cheaper, Steiner-Jovic didn’t say, though he described the cost as a fraction of what would be required for a full-blown DA deployment.

It’s an interesting approach to a problem that utilities are only just beginning to tackle. The World Bank estimates that $202 billion worth of electricity generated is lost per year due to technical failures or theft. Some emerging economies, such as India and Brazil, can see those losses rise as high as 30 percent to 40 percent -- but North America, which has a minimal power theft problem in comparison, still loses about $6 billion a year to “energy diversion,” according to World Bank figures.

British Columbia in particular suffers from power stolen to fuel illegal marijuana grow operations, with estimates of the resulting power loss ranging from $12 million in 2003 to $100 million in 2010. Interestingly, one key problem in proving such theft figures is the lack of data of the auditable quality that Awesense’s system delivers -- making the smart grid a tool in courts of law, as well as in the utility control room.

Not surprisingly, however, Steiner-Jovic declined to comment on whether any of Awesense’s clients were using its capabilities to bust power-stealing pot grows or other illegal operations -- after all, thieves are on the lookout for utility workers who might be spying on their power use.