During the past two decades in many parts of the world, smart meters have been the tool of choice to tackle the problem of non-technical losses often caused by theft of electricity. Despite the installation of over 600 million such devices globally, the value of avoidable losses continues to increase; they still represent between 80 and 100 billion dollars of unrealized revenue annually — and the problem remains largely unresolved.
Additionally, eliminating all non-technical losses would reduce carbon dioxide emissions by as much as 550 Megatonnes each year. This would have the same effect as shutting down more than 100 coal-fired electricity generating plants.
Ongoing efforts to reduce losses may be supported using smart meters, but many of the avoidable losses due to theft, unbalanced loads, wiring errors, metering errors, and other inefficiencies can be difficult to pinpoint and quantify using smart meters alone. Many utilities have no way of measuring localized losses and can only guess based on the difference between estimated load levels on transformers and feeders, and the total energy sold. Further, smart meters are ineffective in identifying losses in the medium voltage grid, an area of particular concern as these losses tend to be the largest and the most dangerous
Only with actual data from within the grid can utilities gain the necessary resolution to pinpoint the location and determine the cause of losses, and make more informed decisions to optimize their grid operations.
This article outlines how utilities can use Awesense’s True Grid Intelligence (TGI) technology to support their smart meter/AMI strategy before, during and after deployment — or without smart meters — to improve efforts to reduce avoidable grid losses and accelerate revenue recovery.
Smart meters around the world
Global smart meter deployment grew by 15% in 2015 with over 622 million meters installed around the world.
Smart meter deployments in the US alone will reach 90 million by 20202. The business case for Smart Meter Infrastructure (SMI) investments often depends on costs recovered due to theft detection. In Canada, for example, BC Hydro’s business case for a $900M smart meter program was supported by $732M of anticipated revenues from their loss recovery program over ten years.
The EU aims to replace at least 80% of electricity meters with smart meters by 2020.
In Latin America, investments in smart metering have lagged somewhat but in some countries that is changing. Mexico, for example, has begun the process of selecting vendors for the eventual rollout of a total of 30.2 million smart meters between 2016-2025.
Smart meters alone are not enough: the advantage of in-grid data and analytics
Although SMI vendors consistently present their products as the solution for eliminating theft and other losses, utilities have discovered that this is not often the case. Despite the heavy investment in SMI, losses remain at stubbornly constant levels in many parts of the world.
There are a few reasons for this gap between expectations and results:
Installing a new meter at the point of service will often uncover illegal connections and wiring or metering errors. In some places, the conductor between transformers and point of service needs to be upgraded and this will also eliminate such legacy problems. But these early wins are not related to the capabilities of the meter — they can be smart or conventional: it is the replacement that reduces loss.
When the meter is communicating and meter reader visits are no longer required, there is a good chance that losses will return over time and may even get worse.
Smart meters create orders of magnitude more data than conventional metering systems, and this demands a significant investment in meter data analytics to identify possible consumption anomalies, tamper flags, etc. Mustering the IT resources, the data science expertise, and the know-how for an effective implementation takes time; there is often a multi-year delay between meter deployments and functioning analytics systems to process this data. In many cases, attaining full system capabilities remains elusive.
Incorrectly installed or incorrectly configured meters can result in losses, even if the meter is smart. Smart meters also do not help when there is loss caused by illegal or unintentional connections between the meter and the supplying transformer or upstream substation. Such unmetered or unknown loads can contribute significantly to overall losses.
Avoidable technical losses caused by unbalanced or heavily loaded transformers and conductors may be estimated using consumption data from smart meters, but if there are unknown and unmetered loads, the smart meter data alone cannot determine the actual severity of the problem.
In contrast, Awesense’s True Grid Intelligence (TGI) technology obtains in-grid data via IoT3 sensors on primary (medium voltage) or secondary (low voltage) conductors between the substation and the billing meter to make it possible to pinpoint the cause of all types of technical and non- technical losses.
Identifying many types of losses.
Whereas smart meter data can sometimes be used to identify diversion from low-voltage lines or meter tampering (Figure 2) TGI can identify these and a range of other non-technical and technical losses: e.g. diversion at the transformer, unmetered loads, illegal connections to medium-voltage lines, unbalanced transformers, overloads, and meter and wiring errors (Figure 3).
How Do Losses Impact Your Utility?
Every year that your utility accepts avoid-able losses as a cost of doing business is another year that you may be:
- missing millions in potential revenue;
- unfairly passing on the cost to honest customers at risk to your reputation;
- putting your field staff and the public at risk;
- reducing transformer lifespan and risking outages;
- making unnecessary power purchases;
- needlessly affecting energy forecasting accuracy.
TGI concurrently identifies those high loss areas that will provide maximum benefit from the immediate installation of smart meters.
Utilities seeking to maximize the return on investment of their smart meter deployment (or indeed any AMI investment) need a system like TGI that provides intelligence on true operating conditions within the grid.
Using energy balances to measure losses
Some utilities will deploy in-grid smart meters at the source of a circuit or feeder (or on the distribution transformer) to target non-technical losse, like theft, by calculating an energy balance between this reference point and the aggregate downstream consumption. This is an effective strategy only if the theft is large enough that it is not lost in the “noise” of the measurement system (i.e. the accuracy — around 1% — of the split core CTs is usually the limiting factor). Too often, however, the cost of permanent in-grid metering is too high, and not enough meters can be installed to achieve the resolution necessary to identify losses.
For example, consider a circuit with 100 consumers, with technical losses estimated at 2% and metering accuracy of 1% . In this situation, theft by a single consumer’s total load might only represent 1% of the total load in the circuit and, therefore, would be indistinguishable from the technical losses and measurement uncertainty. Conversely, if there are only 10 consumers, then one consumer’s load represents 10% of the total and is likely to stand out. A feeder with only 10 (residential) customers would be unusual.
The conclusion is that smart meters alone, or simple feeder monitoring alone, cannot cost effectively provide the level of resolution necessary to identify and reduce avoidable technical and non-technical losses. With roughly 70 million kilometers of power lines worldwide, the only realistic way to know what is happening on the grid is to intelligently sample and apply specific in-grid analytics.
Data Analytics and Risk Modeling
How TGI Works
TGI uses a representation of the distribution grid based on manageable segments that, generally, can be investigated by a typical field crew in a normal working day. TGI then uses a risk model to determine the risk of losses in each segment. The risk models are developed using a range of utility data sources, as well as data obtained from the TGI sensors. Different risk models will reflect the different business drivers: e.g. reducing theft, reducing transformer losses, quantifying losses due to a targeted type of consumer, and so on.
Based on the risk scores, TGI then generates work orders for recommended placements of in-grid sensors and other field tasks. The system pinpoints the most likely sites to conduct investigations while providing the case management, analytical, and financial management tools necessary to bring these investigations to an appropriate conclusion. TGI learns as it goes – so that each subsequent application of the TGI methodology gets more accurate at predicting risk.
Augmenting Smart Meter Success
Analytics and in-grid data can help utilities at every stage of AMI/SMI deployment; from just starting to fully implemented, TGI can help to identify and to correct avoidable grid losses.
Scenario 1 — We Want to Deploy Smart Meters, and We Need to Build a Business Case
Utilities are aware of the many merits of smart meters, but they typically must build a business case for regulatory approval. This is particularly true when seeking a rate increase to support the program. Most such business cases lean heavily on two value propositions:
- Efficiencies gained from not having to send field workers to each premise;
- The promise of recovered revenue from theft and other losses.
These are potent arguments in favour of smart meters, but because deploying smart meters requires a major capital cost, it may be years before the return on investment is realized. Regulatory bodies and other authorities may need additional convincing.
For a stronger business case, utilities can recognize and describe the synergies between smart meters and TGI. Utilities can use TGI to audit their existing power distribution system and obtain a realistic estimate of the types and financial value of losses present in a target area, along with the likelihood of recovery.
In many cases, utilities have a woefully inadequate understanding of the losses on their grids. An audit allows them to demonstrate the anticipated revenue recoverable from loss reduction and helps them build a compelling business case they can back up with evidence.
Scenario 2 — We’re Rolling Out Smart Meters and Looking for a Planning Tool
Due to the significant capital investment required, many utilities deploy smart meters in a staged rollout (for example, 10 percent of the grid per year for 10 years). And because of the costs and the potential for rate increases, these projects are generally subject to intense public scrutiny. Utilities can benefit from knowing where to start — which area will offer the best, most demonstrable return on investment right away. The financial risk and possible reputational risk associated with a massive capital project can also be mitigated or hedged by having a complementary and more agile system that can deliver results faster.
A utility may decide simply to start in the areas with easy access or in a particular management zone, or they may target industrial/commercial clients first. But these approaches may not yield the best possible (or fastest) return on investment. For better immediate results utilities can use TGI to identify the areas of their grid at highest risk for loss. Targeting the riskiest areas is more likely to yield recovered revenue faster.
Other planning uses
In-grid data and analytics can also be used to plan asset spending generally. For example, some utilities use TGI to determine which transformers are running hot and are at risk of failing. Installing permanent sensors at every transformer is not economically feasible in most jurisdictions. TGI allows utilities to sample data from strategic locations and build a model that reveals true grid operating conditions.
Scenario 3 — Smart Meters Installed, Losses Still High — What’s Going On?
Utilities around the world that are counting on smart meters to help reduce losses are often faced with a gap between expectations and reality.
Many utilities find their SMI:
- Generates 80 to 90 percent false positives causing a high volume of fruitless and expensive work orders for unnecessary meter inspections;
- Eliminates the need for meter readers so utilities lose their “eyes on the street” and unauthorized connections go unnoticed;
- Requires big data analytic solutions to process terabytes of data and generate many thousands of additional alerts requiring further processing.
The last point is particularly important because many utilities lack the IT resources and support to interpret the vast quantity of data. And even if they do, most analytics solutions provide field crews with little better than a list of thousands of addresses. With few clues to prioritize their work appropriately and strategically, the efficiency of field investigations is often quite low.
Responding to regulatory pressure
An audit can demonstrate to regulators diligence and a commitment to proac-
tively combating losses.
Conversely, by using TGI to assess risk in selected areas of the grid, augmenting the results of SMI, it is easier to detect the difficult-to-find losses and thus increase the overall value of the smart meter deployment.
Scenario 4 — We Can’t Afford Smart Meters and Never Will
Many countries in the developing world and, indeed many developed nations, simply cannot afford smart meters for every customer. Frequently, the payback period is too long: it may cost $200 for a meter to measure a $20 monthly bill, and the investment cannot be justified.
TGI offers the advantage that it can work with or without smart meters, whereas other big data analytics platforms are dependent on smart meter data. Because TGI samples data from directly within the grid using mobile IoT sensors, TGI can collect data otherwise unavailable. This in-grid data provides a better understanding of the operational conditions inside the grid - including the detection of missing energy.
TGI technology is also scalable and has a modest IT footprint since it runs in the cloud. Utilities with limited resources can operate a “fit-to-budget” loss reduction program using TGI and then use the resulting revenue to reinvest at a pace that works for their business.
Utilities today cannot continue to let millions of dollars worth of energy disappear year after year. Regulators exert steady pressure to make demonstrable efforts to reduce losses and to increase efficiency, rather than continuing to pass the cost of losses on to consumers.
In the US, the atmosphere for approving rate increase applications has chilled; in 2015 nearly three-quarters of the requests to increase fixed fees were rejected outright or modified to be smaller, incremental increases. Environmental motivations are also gaining traction as consumers and society in general seek to minimize unnecessary greenhouse gases. Saftey is also a powerful driver for utilties to address electricity theft. Undetected illegal connections can lead to fires, injury or electrocution. posing a threat to both worker and public safety.
Although smart meters are a desirable part of creating a smart grid, they are tremendously expensive to deploy on a large scale. Unless augmented by an in-grid monitoring capability like TGI, smart metering systems alone are not guaranteed to deliver the loss reductions that utilities, their regulators, or their customers expect.
Awesense’s TGI can complement smart meters. Utilities can capitalize on the benefits of a smart meter deployment by including TGI as a complementary loss mitigation system that increases the effectiveness of the costly smart meter network.
Awesense’s TGI can stand on its own. In some cases, there is simply no business case that can justify the expense of smart meters. But utilities can still pursue a proactive program for loss reduction using TGI. Then, with the added revenue from reducing non-technical losses, they can modernize selected parts of the grid and systematically deploy smart meters as the additional revenue provides funds to do so.