A dozen water utility AMI best practices


By Carolyn Kinsman

To appreciate the value of utility best practices, the stage must be set with some simple definitions and assumptions. It is important to note that the definitions below pertain to water AMI and do not necessarily fit when applied to the electric AMI industry.

AMR – automated meter reading – refers to systems where the meter reader must be in close proximity to the water meter and acquire a remote meter read using a handheld receiver or by driving a vehicle down the street that accesses the reads using a mobile data collector. Typically this type of system delivers a cumulative read when it is interrogated or receives the read.

AMI – advanced meter infrastructure – refers to systems that enable meter reads to be transmitted over a fixed communication network architecture to a data collection computer housed at the utility offices. The AMI system may be one way or two way and can provide a minimum of four reads per day from every meter to the AMI data collection and control computer.

A few fundamental assumptions about the AMI system features also assist in building out the reasoning for utility practices. For this article we will assume that:

  1. The battery life of the AMI module is 20 years
  2. Most technology related investments (computers, networks, CIS, etc.) are lucky to last 20 years
  3. The water utility deploying AMI requires a minimum of four reads per day (or more) from a meter.

AMI best practices for water utilities have evolved over the past 10 years. This is approximately when the first water AMI fixed networks were deployed to read water meters. Before that there was only AMR. The utility trailblazers who implemented these inaugural systems had little to no guidelines on how to configure and deploy the network and how and what to do with the exponential amount of data that they could now collect. AMI selection in many cases was based on the faith that more data from the meter really did have intrinsic value to utility operations and by default it would also provide customer benefits as well. The early experiences from these trailblazing water utilities are now what assist others in avoiding the same painful pitfalls they encountered. Hindsight makes some of the early oversights almost humorous. However, the utility that has to live with some of those repercussions may not find as much humour in the situation as those of us with armchair wisdom.

If one is just starting to consider AMI or is getting ready to deploy a system soon, some of the best practices may smooth the road of a few speed bumps. The following is a brief list of some of the more generic lessons learned. Hopefully these will assist in looking more closely at some of the technology assumptions that may currently be taken for granted.

If the water meters are in pits, it will be necessary to change out the iron and or cement lids for polymer. Depending on the vendor one may also have the option of drilling a hole in the lid and installing an antenna. Regardless of when an AMR or AMI system is going to be considered by utilities that are in this situation, a change in current public work specs to move to polymer is a good idea. Polymer optimises the ability of the RF signal to transmit out of the meter pit. The cost to replace these during the AMI deployment just adds to the capital investment that has to be cost justified. There is no time like the present to start getting ready for the day when AMR or AMI will be deployed.

If the meters are located in basements, one must reach a pivotal decision on where to install the AMI module. Some utilities think an external mount enables maintenance and repair to occur without access issues after deployment. Still others look at historical resilience of the AMI module in the field and opt for an in-home installation which alleviates costly and extensive wiring while maybe even enhancing AMI battery life.

If a utility has outdoor touch read units they may think it is cost effective to use the existing wiring to connect to the AMI module. The reality for most utilities is that the touch read is a two wire connection and they will need a three wire connection for the AMI module. This means the utility will have to gain access to the meter to make this connection. Because of this utilities may not wish to factor in existing wiring in the decision made in relation to Best Practice #2. A good practice for those considering their existing wiring to be a benefit is to first validate the situation with a survey of the current connections.

Regardless of the decision to consider AMR or AMI a utility has to know what type of meters it has deployed. In most cases utilities have a pretty good idea but not an exact one. AMI deployments are not like playing horseshoes or using hand grenades – close is not good enough. An exact inventory is necessary.

Generally speaking, residential water meters have an acknowledged accuracy life of about 20 years. For this reason some utilities may consider keeping meters that are under seven years old. However, if this decision is made the cost analysis should factor in what repercussions could occur when keeping some meters out in the field for 27 years.

Some utilities may disagree with the established average of a 20- year meter and register. However, they must also weigh current policy with the perspective that the value of the AMI in many ways is basically the elimination of virtually all visits to the meter unless there is a known problem. This means it could be 20 years before a site visit is required to deal with the AMI battery replacement. Best practice at this point is to replace the entire AMI module due to technology obsolescence. This being the case the utility may also want to reconsider how it manages meter life cycle replacement policies.

Should one be a utility that has exemplary water and hence exemplary meters, and registers, one must also consider that by keeping a percentage of the existing meter stock one will be breaking up the mass AMI installation process by separating meters into a retrofit category rather than everything being a straight change out. This requires a different process that may prove to be more costly than just contracting for a complete change out process. Utilities considering this should take into consideration whether these relatively new meters are dispersed throughout their service territory and whether the retrofit will result in more trouble and costs than it is worth.

Complete change out of the meter and purchase of the AMI module allows the utility to deploy factory potted (no in-line connectors) units. This provides the ultimate level of water tight connections and has been proven to alleviate a greater percentage of failed units in the field due to water seepage issues.

Most computer-related technologies are not expected to last 20 years. It isn’t that the technology stops working; it is just that the units become obsolete. For this very reason utility business cases must cost justify the purchase of the AMI over a significantly shorter period of time than 20 years. As AMI is a technology decision most closely rooted to telecommunications and IT expenditures, the business case should adhere to the expected longevity of this type of investment. Rule of thumb: The payback for the AMI including new meters and installation should not exceed 8 – 10 years.

As utilities have already experienced with handheld reading devices, the technology must be upgraded every 8 – 10 years. All technology devices are dependent on the components and RF modules being sold by the technology component manufacturers, who in turn supply the AMI vendors. When the component is no longer offered then the AMI vendor is notified of the pending termination of supply. The AMI vendor retools the unit and produces the next module version. At the same time, the AMI vendor in turn notifies the customer about the termination date of the current production module and the date when support will no longer be offered. This is another reason to make sure cost justification does not exceed 8 – 10 years.

For the above reasons water utilities should carefully review their previous business practices for meter maintenance and depreciation. With an AMI module now in the picture, utilities must develop a business practice that considers all these elements in the overall AMI evolution strategy moving forward.

The AMI system selected may be one-way or two-way but must (if one has a fixed network) at least be able to alert the AMI system administrator that leaks or low flow is occurring. For this type of reporting the meter must transmit reads at a minimum frequency of four times per day – ideally every six hours.

What most utilities fail to realise is that even though the AMI system can read four times per day or maybe even hourly, it can only read what the meter register is able to relay through the AMI system. Many utilities have been very surprised when the AMI system they purchased produces four reads in a day but the read does not change because the register can only transmit in 50 gallon increments. This is another good reason to do some homework on meters before purchasing the AMI system.

After the AMI system is deployed the idea is never to visit these meters again unless it is to perform maintenance. The longer the AMI is out there the more in-house knowledge about where the meter is located disappears. For this reason utilities should stipulate that the installation contractor acquires the meters’ GPS locations and that these coordinates are put into the GIS.

While not as high tech a solution as the previous 11, many utilities also find it good form to paint a line on the curb at 180 degrees to where the pit is located. This visual aid acts as a great backup measure for finding the meter once the meter readers are no longer regularly visiting these meters.

There are more than this dozen of best practices. What this article is meant to show is how much “new” thinking must occur with the deployment of the AMI. The good news is that more and more utilities are choosing to deploy and with this, the map for instituting a smooth transition with no surprises has greatly improved. Hopefully when you have completed your AMI deployment you will have a few best practices to share with the rest of us.