By Patricia Thompson and Michael Ozog
Over the past year or two, many utilities have begun to formulate and execute smart grid strategies. The bulk of these strategies depend on the installation of smart meters. Although some argue against the necessity of smart meters to execute these applications, the industry appears to accept that more and better energy usage data provided to customers will lead to improved energy management, an increase in satisfied customers, and more efficient utility operations and planning.
However, the ways by which these smart applications will use the new streams of smart meter data is not obvious. Here, we lay out examples of how some utilities have begun to use smart meter data to improve customer focus, expand targeted energy applications, and enhance operational planning – in essence, enabling the “smarts” in smart grid.
CUSTOMERS RECEIVE ADDITIONAL BENEFITS AS UTILITIES MITIGATE COSTS
Because widespread smart meter installation is costly, and the benefits from improved operations often do not fully cover the costs, utilities often opt to apply smart meter data to improve customer service and energy applications. Such benefits might include enhanced pricing options, new energy efficiency programmes, electric vehicle charging options, and enhanced billing options. New benefits result from smart meter infrastructure, mitigating the infrastructure cost burden – one that is likely to grow large enough to warrant increased regulatory and consumer scrutiny. The continued lack of physical inventory buffers between supply and demand continues to pose financial risk to utilities and consumers alike. Where supply is short, weather extreme, or new loads emerge on the horizon (e.g., PCs, plasma TVs, DVRs, electric vehicles), utilities face increasing peak load risk. Because electricity cannot be stored for a reasonable cost, peak load costs can sometimes skyrocket.
For example, some regions have historically experienced US$1,000 to $2,000 per MWh or higher under extreme conditions, which is many times the average cost of power to customers – clearly an area of concern for utilities and consumers alike. Even if a utility currently enjoys flat loads, supply mixes could be jeopardised by the advance of electric vehicle peak time charging, increasing electrification, national security interests that seek to avoid oil, and other circumstances.
WAYS UTILITIES CAN USE SMART METER DATA TO SERVE CUSTOMERS
More accurately quantify the costs to serve.
Many utilities are surprised to see just how different the (economic) cost to serve varying customers can be. Street lighting (night loads) can be three to five times less costly than a customer operating more on peak. Similarly, the difference in costs between serving an electric vehicle smartly charged at night and the same vehicle charged on peak is significantly. In the zeal toward smart meter installations, utilities should plan ahead and know which customer loads are more likely to be cherry picked by competitive threats. Current consumer perception is that oil prices determine fuelling costs. But relative to how cheap electric fuelling costs can be, our research suggests that many if not most consumers may ignore traditional peak hour prices, opting for convenience over cost savings.
Smart meter data can aid in long term distribution planning.
Extending the electric vehicle example, smart meter data may be necessary to mitigate the long range risks of increased electrification, customer choice and aging distribution infrastructure. Consider secondary service transformer overloading risks. Such burdens generally stem from increased electric load growth (e.g., electric vehicles) and can include large additions (perhaps with fast-charging capabilities, high KVA needs). Conceptually, it is easy to see that aggregating smart meter data from the homes, by hour or minute, allows the utility to clearly determine if or when adding fast-charging electric vehicles to that secondary service transformer matters, or not, or if it should be changed out in advance.
Generally, the risk to the distribution system from electric vehicles tends to diversify as we move up the circuit toward the substation. However, because many customers in close proximity tend to behave similarly, there is still significant localised risk to secondary service transformers. Imagine a neighbourhood of car enthusiasts who watch their neighbours drive a sexy new Fiskar Karma or Tesla down the street. Human nature suggests that it won’t be long before more of these vehicles appear on the street, posing a local risk to that section of circuit. Our research over the past two years has clearly shown that “birds of a feather flock together.” For utilities, this means distribution risk, which can be uniquely mitigated via access to hourly, minute-by-minute access to the smart meter data for each home. Interestingly, our research also suggests that those who appear most likely to charge on peak (moms and mini-van drivers) also appear to be the least price sensitive to financial penalties on peak.
Utilities are starting to gauge risk management and the value of energy operations and customer offers.
This is done through two-way signalling, either informed by smart meter data or communicated via smart meter equipment. Here, many operational benefits are achievable via broadband, or relatively inexpensive relay devices in the home, or affordable current transducer measurement devices – none of which requires a utility to install a smart meter directly. Broadband communications and the recent advance of affordable relay and signalling devices create an option for utilities that desire low cost options for customer-focused offerings, especially if the utility is willing to forgo or unable to realise the operational and labour savings ascribed directly to the meter itself. In other words, if the utility does not care about remote disconnects, time-of-use pricing or revenue grade usage observations (i.e., if plus or minus 3% is acceptable),smart grid applications without the use of a smart meter may be effective.
The operational applications and simulations shown here can be enabled either way. Most important is the manner in which a utility starts thinking about managing load changes. With direct observation and management of customer energy use enabled via two-way signalling, we show how the direct manipulation of the loads themselves open opportunities for consumers and utilities alike.
An inventory buffer can smooth operations while increasing customer satisfaction
We have created a solution for a critical missing element of utility operations: an inventory buffer, whose implementation actually increases customer satisfaction through customised billing options and greater control over their loads and costs. Moreover, we have shown via simulation, using sample results from US pilot programmes, full control over the peak load risk with only 30% to 40% customer participation. It is unrealistic to expect 100% of customers to participate in utility enabled energy management applications. And it might be cheaper, to use an already installed broadband network linked to existing meters, and only approach 30% of the customer base.
INTEGRAL ANALYTICS PROVIDES PROOF AND RESULTS
Value customers, one at a time
Here is an example comparison of the relative cost to serve individual customers compared with an average class rate. Some customers have more cost management opportunity than others (Figure 1).
Clearly, new entrants to this market who are offering bundled energy services are likely to target the less expensive loads if the regulator forces settlement of the energy against the class average. Over time, this leaves the utility with only peakier, more costly customers, which inevitably leads to higher rates for the customers remaining with the utility in that class. This further exacerbates the ability of the new entrant to cherry pick inexpensive loads year by year. As smart meters roll out, sooner or later, competitive marketers will ask to settle on this versus a class average.
Zero In on which customers to target
With smart meter data, costly customers are quickly identified. Utilities can manage cost savings via precise targeting of energy efficiency measures which meet their needs. Figure 2 shows an example of a neighbourhood where the cost to serve translates into the type of energy efficiency measure that best reduces both utility and customer costs. Through analytic modelling of the smart meter data, we discern which homes are inefficient. Now a utility can more cost effectively target an appropriate rebate, offer or pricing programme to the right customer segments.
This same set of analytics also helps the utility identify the ceiling on achievable avoided costs. Many homes are already efficient, and targeting them with energy efficiency measures repeatedly is not likely to yield incremental avoided costs gains, relative to the marketing costs.
Expect distribution impacts from electric vehicle adoption
By combining hourly patterns of home usage by customer type with marketing research forecasts of customer segments that are likely to adopt and charge different types of electric vehicles, we can begin to forecast the long term risks, consequences, and value from more widespread adoption of these vehicles. Figure 3 shows that future growth of electric vehicle loads may emerge within pockets of more rural areas, but generally will be more preferred within five miles of an urban centre. Given the likelihood that distribution infrastructure is more aged in the region depicted, advanced long term planning with regard to substations, circuits, capital budgets and utility sponsored charging facilities is warranted.
Identify at-risk transformers in the long term
To be clear, the market research forecasts of future adoption are not reliable enough to pinpoint or predict the future behaviours of any customer or any one home. However, as repeating patterns of forecasted growth across a neighbourhood arise, especially where aging infrastructure may pose additional risk, the analysis does suggest that distribution planners should begin to consider risk mitigation strategies within some high risk circuit sections over the next few years. The map in Figure 4 depicts a set of load forecasts relative to the service transformer ratings, across 1% to 20% simulated market penetrations of electric vehicles.
Interestingly in this case, two circuits reveal that the same set of 10 to 20 transformers are at risk whether electric vehicle adoption increases 5% or 20%. Either these transformers were likely undersized to begin with, or the families that have moved in have increased the loads well beyond normal expectations.
Here, smart meter data analytics identified exactly which transformers are at risk in the long term, which can save the utility significant time, effort and capital costs if they have to retrofit the neighbourhood over the next 10 to 15 years, given average forecasts of electric vehicle adoption.
Manage risk through operational programmes with customers
Figure 5 depict results from US pilot programmes in which twoway, broadband communications allow customers to control their appliances, set temperatures, adjust lights, and participate in utility sponsored billing options which are mutually beneficial. The clear flattening of loads mitigates peak load risk. The slight bumps in the load are caused in part by dual or multiple objectives set by the particular utility to favour cost savings over purely flat load.
In the pilot results, no customer experienced any change in comfort, control or usage. By giving customers more options to control their own use and partnering with customers in unobtrusive ways (such as customers having the option to override at any time, but none did), the utility began to create a virtual buffer between supply and demand. In effect, this serves as a real-time inventory, mitigating the risk of peak supply costs.
Not all customers care to engage with a utility, but of those that did take action, energy savings averaged 8% to 10%. Further, peak load reductions of 10% to 15% or more were realised without the customer’s comfort, convenience or usage being compromised.
Prepare for peaks with EV adoption by managing the load operationally
The promising future revealed from smart grid pilot results extends to an example circuit simulation. This demonstration offers two key insights (Figure 6). First, increased electric vehicle adoption may create new system peaks, where customers are not dissuaded from peak-time charging. This poses new risks to otherwise stable supply mixes for some utilities. Second, peak load risk can be managed operationally using a fraction of the total customer base thanks to new two-way enabled systems, such as smart meter, broadband, cellular and more.
In this example, a 40% participation in utility sponsored load shifting on vehicles, water heaters and air conditioners resulted in eliminated peak load risk, lower total generation costs, and increased customer satisfaction. This includes customers’ ability to control their own usage and bills (Source: Integral analytics IDROP software simulation).
Additional supply side benefits emerge from the creation and management of this type of balanced and optimised dispatching which are beyond the scope of this article. They include participation in ancillary service markets, “cloud following” in the case of solar, “plant following” to achieve preferred operations for thermal units, “wind following,” and frequency or reliability benefits, to mention a few key benefits.
Offer unique billing services, including bill certainty
Utilities can begin to offer more unique billing services, which give customers the ability to exactly specify their desired amount for their bills or lock in a guaranteed percent of conservation or energy savings.
Because customers set their preferences, but can override them at any time, satisfaction increases. Moreover, the utility can now offer bill certainty instead of just increased control over usage. The increased appeal of bill certainty means that utilities have a much better handle on cost management. In some cases, the avoided costs to be gained can double and triple with lockedin customer conservation goals of only 3% to 5% savings.