March 29, 2024

Smart Grid Calling: Do You Know Where Your Outages Are Tonight?

by Charles H. (Chuck) Drinnan, President & Principal Consultant, eWAM Associates
It's midnight and several meters are experiencing power interruptions and have sent last gasp messages over your new Advanced Metering Infrastructure (AMI) communications network. You promised you would improve outage response times for your customers who invested in Smart Meters. So now it is midnight… do you know where your assets are – and how to repair them?

Advanced Metering Infrastructure (AMI) offers utilities new measuring devices – Smart Meters – that revolutionize outage identification and response, reduce system Customer Average Interruption Duration Index (CAIDI) and eliminate unnecessary (and expensive) truck rolls to verify outages. If you know where your facilities are located and how they are connected using an accurate network model maintained by your Geographic Information System (GIS) and are willing to integrate AMI and GIS with your Outage Management System (OMS) and Interactive Voice Response (IVR), you can reduce costs and improve customer service and become a leader in the Smart Grid revolution.

Determining Outage Validity and Primary Outages

Utilities typically implement AMI-assisted capabilities in stages so that each stage demonstrates improvements and leads to the confidence to implement the next stage. The first stage uses AMI pinging capability to determine if power has really been interrupted. That is, when a customer calls in reporting a suspected outage, the AMI is used to ping the customer to see if the power has actually been interrupted. This eliminates expensive truck rolls to determine whether an outage has occurred.

The next stage automates this process and supports batch pinging operations. So, if a number of outages are reported, the OMS predicts the control device that has been tripped, pings all of the customers that should be affected by the outage, defines the outage as a primary outage, and dispatches the appropriate crew with the right equipment to the suspected trouble location. During normal hours, the IVR contacts the customers who are out of power and tells them the utility knows they are without power and is doing its best to restore power as soon as possible.

Automatically Reporting Outages

The steps described above use AMI to validate outage situations. The next stage uses the Smart Meters themselves to report an outage. When a Smart Meter senses that it is losing power it sends a “last gasp” message back to the central office. This message is interpreted as a new outage in the OMS even though the customer probably hasn’t called yet. Typically the IVR tries to communicate with the customer advising the status and expected repair times. In many cases, the system may identify outages and dispatch crews to repair them before the customer even knows that power has been interrupted.

Managing Multiple Masked Outages

The next stage uses the Smart Meter capability to sense that power has been restored. During a storm there can be – and often are – multiple outages situations affecting a given customer. Some of these outages are masked by outages further up the line. If the first outage occurs ‘down-line’ (i.e., logically near the meter), the system may be able to identify the outage as an individual outage even if subsequent outages occur up-line. This wasn’t practical when customers haphazardly reported outages and there was no real ability to pinpoint outage locations easily.

However, if an up-line outage occurs first, the down-line outage is masked from the outage system. But then, when power is restored, each Smart Meter sends a message that power has been restored. This process is used to determine if there are masked multiple outages without the crews having to inspect the entire circuit, end to end. An alternative approach yielding essentially the same result uses the pinging process (described above) to determine if the power has been restored at a specific meter.

Recording and Keeping Track of Outages

Most outage systems record the initial report of an outage and then each operation required and performed to correct it. They record outage duration, equipment repair and replacement, switching operations, necessary follow up, and other activities that affect customer service. Among other things, this information is then used to compute CAIDI and other reliability indices.

Improving the Network

Many utilities either have an electric network model or may be planning to develop one. However, these models are notoriously inaccurate at the specific meter connection level. Some utilities do not model the individual phases down to the customer level, and others don’t have an explicit connection between the customer and the transformer. The costs to define a new network model or repair an existing model through a field inventory are often prohibitive. This is usually because existing map records are poor and field verification is very expensive.

If your network model needs work but has the basic connectivity, you can improve the model using AMI technology and outage events. Every time an outage event occurs a record of the affected facility and expected effect it has on the down stream meters is stored. For example, if a tree causes the failure of a conductor the outage/AMI system can determine and record every meter affected. Thus, the system knows every down line meter on that circuit and can automatically identify connectivity errors quickly, easily and at minimum cost.

The system can also compare the affected meters to the expected outages. If there are meters that recorded failures but were not part of the logical network, you can change the network so that they are included. Typically that is changing the setting of a tie switch. If two circuits are near each other and one circuit fails, move the unaffected meters to the neighboring circuit. If the occurrences and exceptions are automated so that the changes are presented to the user in a well-defined manner, the changes in the network can be accomplished quickly.

Does your network become a reliable network model using this approach? Clearly, if there are many misconnected meters, this process will not produce a good model quickly. A concerted effort results in an increasingly more accurate network model. Some utilities record confidence levels to indicate which portions of the network are accurate and current. But even an inaccurate model will provide the crews with a starting point. When an outage occurs in a low confidence area, the crews are tasked with additional field verification of the connectivity, and the back offce makes a special effort to record the network connectivity determined by the crews. In areas where the network model is totally unacceptable, a field inventory is warranted.

But We Don’t Even Know Where Our  Assets Are!

Some utilities still don’t have any digital network models. Instead, they still do their work by paper-based maps and map books. Or, they may have a series of digital maps, but nothing is connected in any type of documented network architecture. In this latter situation, the utility should implement a plan to develop a digital network model as a fundamental starting point. An incremental approach is advised, and starts with finding a reliable base map to work from. If nothing else is available, start from a Google map.

Next, locate all the meters on the digital map – street address models using service addresses are a good start. Maintain these maps as new meters are added. Determine the growth areas and develop a digital network model placed on top of the land base. High growth areas can often easily justify the cost of defining a digital model. It is generally advisable to complete one circuit – or better yet one substation – at a time. Slow growth areas can be filled in gradually as the opportunity arises.

Assuming the utility has an AMI, when an outage is reported (by telephone) check it using the pinging process and record it on the meter map. If there are multiple outages you will get a good idea where to start looking for the primary outage. Record all the down line facilities, but recognize that they are only a start to a comprehensive circuit connectivity model.

What Are Those Smart Meters Really Doing?

Most utilities implementing AMI are still in the process of installing their Smart Meters. Many of the utilities have plans or contracts to implement AMI-enabled outage management systems after they complete all, or at least a large portion of their AMI installations. Many of these utilities have justified their Smart Grid projects at least in part on improved network reliability and reduced outage durations.

Several utilities have already achieved reduced truck rolls and improved CAIDI results, primarily from incorporating pinging processes into their existing outage procedures. These utilities and the others that have contractual obligations to do so are adding this integration across AMI, OMS, and IVR. They are confident that they will achieve more benefits as they incorporate automatic outage recording from AMI ‘last gasp’ messages. There is still much to do before AMI systems are completed and integration between AMI and OMS is all in place. However, when they do, this new level of integration will become utility best practice – wait and see.

About the Author

Chuck Drinnan is an independent market consultant, analyst and advisor. As President and Principal Consul­tant for eWAM Associates, he brings more than 35 years of utility Transmission & Distribution systems experience and provides consulting and project management work of the highest quality and integrity based on industry best practices. He has experience in all phases of information systems design, from developing and managing one of the first commercially successful GIS databases for the utility industry to specifying, designing and managing the imple­mentation of one of the industry’s most comprehensive enterprise gas and electric systems. Chuck is a co-author of the international Spatial Data Transfer Standard, has organized and presented the GITA (Geospatial Information Technology Association) Work & Asset Management Sem­inar, has authored over 30 technical papers and serves on the Editorial Board of GeoWorld magazine. Chuck can be reached in Houston, Texas via email: chuck.drinnan@ewam-associates.com or by phone: +1.713.461.2606.