Local Energy Networks
Making energy networks legible and user-centered in emerging markets
Our team's intimacy with electricity access issues from working in the Haitian and Kosovo energy sectors impassioned us to create a user-centered system for these markets. We intrepidly (and naively) ventured to solve a plethora of problems:
Utilities lose $ billions to theft and non-payment. Manual processes like reading analog meters, delivering bills and collecting cash payments belabor utility operations and most significantly, there is no effective way to monitor and cull theft. Electricité d'Haïti, the state-owned utility in Haiti, for example, loses over 50% of their energy distribution to theft and non-payment and requires an annual subsidy of $170M to maintain its operations. The utility additionally offsets their losses on existing customers, imposing the highest tariff rates in the Western Hemisphere.
Consumer demand is in the dark. A massive market is willing to pay for electricity but is marginalized by fruitless electricity connection applications, post-paid programs, manual billing and payment centers. As a result, many improvise their connections [illegally] to the grid, causing frequent brownouts and blackouts with overwhelming infrastructure demand.
During 2011 and 2012, we envisioned and engineered new smart meters that remotely connected electricity sensors with grid operators and end-users via cell signal. This ecosystem made pay-as-you-go electricity service possible, opening up a huge market to utilities.
My role: In parallel to managing LET's smart meter engineering efforts as product manager from August to December, 2012, I spent part of my time designing a new interface for grid operators to understand and manage their networks effectively.
Our sensing technology gave utilities entirely new data streams from their networks and consumers, including theft, power usage, messages and payments.
LET's energy system received data remotely from end-user and meter communications, empowering utilities to understand their networks in ways that had been fairly impossible before, especially in non-formal cities where electricity connections are often improvised.
How could grid operators interact with this new intelligent landscape?
Contextual interviews with electricity grid operators were difficult in our early stage. Similarly difficult was competitive analysis on their power distribution and demand side management systems.
I explored data visualization possibilities based on the new opportunities of our sensors and made assumptions about what decisions our users would be making. The most important action our system gave prospective utility users was the ability to instantly start and stop power service.
I initially designed a geospatial interface, iconography and color visualization modes to answer critical questions a utility operator might ask:
- Who is using the most power?
- Who's power usage is the most vital?
- Who is paying and not paying for power?
We organized energy node subnetworks to make energy community-centric and locally controlled.
Our IP and central technology premise was that energy (and its theft) is more effectively managed on a subnetwork by subnetwork basis. Field research and literature reviews revealed how influential social reputation was within communities:
"They steal electricity until the utility decides to “name and shame” the customer, because the cost of social condemnation far exceeds the savings from not paying fully for electricity consumption."
"They stop stealing once they become aware that the utility has the means to detect and record it. Recent experience in such countries as the Dominican Republic and Honduras shows that consumers stop stealing if they face the risk of social condemnation."
-- "Reducing Technical and Non-Technical Losses in the Power Sector", The World Bank Group, 2009
The LET system virtually re-organized the electricity grid into local electricity networks, where community-based decisions could be made. Our technology sensed power usage at the transformer and end user points, allowing us to calculate the magnitude of theft and incent local stakeholders to take action.
How can utilities manage thousands of homes?
I critiqued my earlier interface design and sketched new, network-based organizational structures for the interface.
I abstracted the structure of a electricity transmission and distribution into a graphic network, stepping away from the earlier location- and iconography-based system. A network-based interface gave grid operators context for energy issues, especially the relationship between generation- and demand-side operations.
Integrating improved visualization modes, I designed a network-based interface.
Browsing the electricity network:
Network visualization modes:
Layered on top of a network understanding, grid operators would need additional data visualizations to make decisions.
Our users' key management activities, such as shedding load and determining dynamic pricing events, depended on understanding power data on a timeline. I designed a timeline visualization to show usage data alongside activities and triggers.