Getting the Best Value Smart Meter for Your Money

Mark England | Feb 15, 2011

Utilities and governments must think ahead when planning a smart meter roll-out -- early obsolescence is an expensive error. Customize and build in future-proofing to stay ahead of the game advises Mark England, CEO, Sentec.

With some utilities well underway with smart electricity meter roll outs and now turning their attention to the gas market, and others just starting to look at options for their first smart electricity meters, metrology design and getting the best for your money is an important focus for those tasked with selecting the best option for meter design.

It's tempting, when planning your smart meter roll-out to go for the simplest and seemingly safest design option. After all, you have complex issues such as geographical implementation and timescales to deal with, so why wouldn't you just pick a solution from a long-standing and reputable meter manufacturer?

But does that get you the best bang for your buck? The term "smart meter" has been used (often incorrectly) to represent a variety of products on a spectrum ranging from a dumb meter with a radio attached for remote reading, to a high-tech, high-cost meter with multiple 2-way communications interfaces, supporting every conceivable measurement and tariff option, many of which may not be considered necessary today. Off-the-shelf meters may sound like a quick and straightforward option, but inevitably incumbent meter manufacturers have been forced to make many compromises in the design in order to make a product that could be used by a wide range of utilities. A good analogy would be buying a computer made to your specifications from a bespoke manufacturer like Dell, rather than picking up a ready-made one from a high street retailer. The price would be fairly similar, but the features are tailored to meet your needs, such as a high speed graphics card for gaming or extra memory for image storage.

Demands on the applications of a smart meter will increase as other technology develops. As such, meters rolled out in the next few years must have the ability to adapt to changes in application that may happen in ten years' time, or utilities will find themselves with meters that need to be completely replaced every few years just to keep up. Utilities investing in smart meters need to take into account the advances made in metrology methodology, production methods and materials, communications technologies, electronic components, firmware and operating systems. For utilities without internal knowledge in this area, the best approach may be to work with an experienced technology development partner to provide the technical insight. Utilities should make sure that the partners they select to help them develop their smart meters have enough specialist knowledge to be able to predict applications that may become necessary in future and future proof the meter accordingly.

It is important to remember that different utilities and countries will have different IT systems in place, radically different geographies, population densities and housing stocks, a variety of communication requirements, and different distribution infrastructures. So instead of looking at what's available now, each utility should be thinking about what are the essential and desirable requirements for their smart meters.

Specification Questions

Some of the more obvious specifications relate to, for example, the quantities to be measured and their accuracy limits, the time-of-use tariff structure, and minimum frequency and reliability of remote reading.

The harder ones relate to ill-defined and evolving requirements -- for example:

  • Will it need to be paired with a home energy monitor, to help engage customers, and what sort of depth and resolution of data will this require?

  • What communications means and protocols will it need to support out of the box, and in the future? Does it need to communicate with other smart meters -- maybe an electricity meter that's already installed, or perhaps one that might be installed later?

  • What types of smart grid functionality will be needed in future -- load shedding or time shifting of smart appliances, or control of the charging of electric vehicles?

  • How will it integrate with future distributed local generation and manage future feed-in tariff changes?

  • It's hard to predict and build in all these diverse requirements today, so which hooks and features for adding them in remotely do we need to include?

A clear understanding of objectives should lead to a finished product that will provide the highest ROI for the utility and strong benefits for consumers. In many cases a bespoke design will provide the best balance between cost and functionality for the meter, and if executed proficiently, will also confer a number of extra benefits such as in-field upgrade ability, ownership of the design, flexibility of manufacture and supply chain control.

One of the key ways to build for the future is to specify a meter with a larger flash memory capacity -- this provides the ability to store firmware images for upgrade purposes, and to record rich, deep data to support future applications such as usage profiling and load disaggregation, for almost no additional cost. Using industry standard processor architectures and hardware interfaces between functional blocks means that, as new improved versions of devices become available from different manufacturers, it is relatively easy to update the design to take advantage of them, and mitigate the risk of component obsolescence. Considering the capability to deliver additional functionality, even if not implemented in the initial firmware, is also likely to extend the useful life of the meter.

Traditional meter manufacturers have built their businesses and manufacturing up around the model of a steady ongoing replacement of product as it reaches the end of its working life. The new waves of smart meter deployment require much larger volumes of meters to be delivered over short periods of time, something that not all traditional manufacturers have been prepared for. This has been evident in North America, where the current rate of electricity meter installations is approximately five times higher than the historical replacement rate. This rapid fluctuation of the supply volume is very familiar to the world of consumer electronics, where products are typically manufactured in high volume by contract manufacturers for relatively short production runs. Capacity can be scaled up and down relatively quickly, and at multiple plants if needed. At the end of the product design life, or a particular wave of installation, the contract manufacturer simply reassigns the production facilities to build different products. By contrast, traditional meter manufacturers who have invested in their own capital-intensive production facilities may struggle to meet these peaks in demand, and have to carry the cost of the line and employees when orders are low -- these costs all have to be passed onto the utility as part of the product price, and might affect the cost or payback of the roll-out because it is extended over a longer period.

For utilities specifying meter design, the ability for the finished product to be built by a contract equipment manufacturer (CEM) is an important part of the design challenge. This ranges from taking advantage of the tremendous buying power of the CEM by using components widely used in other high volume products, making best use of the CEM's PCB manufacturing and test capabilities, designing out steps or processes with high capital equipment or fixturing costs, using commonly available materials and processes wherever possible, and avoiding single-sourced or long lead-time parts.

Decisions made by utilities now, and how they choose to spend their money, will have a big impact on the success of smart meter roll outs and how consumers perceive the technology in the future. Utilities are in a position to decide on the requirements for their meter, and produce the exact instrument they need, and ensure it will be a valuable asset for a suitably long period. It is well worth investing up front to provide the best ROI for the roll-out.

Commissioning a bespoke design means the resulting product can provide a truly smart meter with much more potential than existing offerings. Above all else, it will match the specific utility requirements, and provide a longevity and upgrade ability unrivaled by off-the-shelf products. It is this upgrade ability that will be key to ensuring the consumer engagement required to make the project a success. The future is in our hands -- it's crucial that we make the right choices.

Related Topics


Very good article that explains the biggest problems with smart meter commercialization. Under the current system of utility-owned meters, these are daunting problems when money to invest in future proofing their smart meter choices is often scarce.

"It is well worth investing up front to provide the best ROI for the roll-out."

This if far easier said than done I'm afraid when utility companies must justify up-front spending, particularly if they must pay for these up-front investments by applying for rate increases or adding specific extra charges to all customers bills. Regulators must approve of these bill increases, and even if they do, the consumer backlash from such bill increases can derail these projects UNLESS all customers can see and feel direct benefits from such spending up front, or be convinced they will see them in future.

The utility companies thus have a public relations challenge on their hands to convince all their customers they will somehow benefit from smart meter investments. This situation will prevail as long as all up-front spending must be paid for uniformly by all customers in the current regulatory environment that most utility companies are forced to operate under.

The only way out of this predicament, in my humble opinion, is for regulation to change that would allow utility companies to raise money in other ways, such as selling optional services to interested customers willing to pay for them, much like our CATV and telephone company business models. But sadly this type of regulatory change isn't even on governments’ or regulators' radar screens.

Your warnings about early obsolescence and future proofing are excellent. Very important.

As evidence of obsolescence ALREADY happening here in Ontario, we now have all 5 million or so distribution utility customers on smart metering from various manufacturers, in particular Elster Metering, Landis+Gyr (with Trilliant Networks technology), and GE to name some of the predominant ones. Most of these can be considered obsolete in some aspect or another because they weren't equipped at the factory with all the potential hardware or software options the meter manufacturers have to offer.

To make matters worse in Canada, any software shortcomings that could technically be fixed with simple firmware reprogramming of the meters are not allowed by our federal agency Measurement Canada. They prohibit field changes of any kind to a deployed meter, so the only way for utility companies to upgrade anything in the field is with total meter replacement with models whose software and hardware configurations have been preapproved by Measurement Canada.

Future proofing is probably impossible from the perspective of decades of attempts in the IT industry to do the same. First, the product production cycle is limited unless fully custom. Mark references Dell as a "bespoke" manufacturer but the truth is that they integrate parts off the market and have them assembled to order. Every day the components change as Dell buys parts on the market. The differences between models are constant even if the model number does not change. This will likely be the pattern for many "smart" devices and not just meters since buyers (utilities) are always price sensitive.

Second, we have already seen wholesale replacements of products in a variety of locations for problems with product failure or software update failure. SDG&E replaced tens of thousands of meters pro-actively when a software update broadcast took thousands of meters offline. A fully custom or off the shelf product would have still needed replacement. As Bob has stated - Canada is unable to even accomodate software updates for meters which basically eliminates 99% of the advances possible without pulling the meter. (OMG!!)

Third, the costs to pull a meter for a repair are almost the same as to replace the meter with a newer model. This is also the case for physical upgrades. If a truck roll is needed to replace an $10 part or install a $20 memory card upgrade - the same truck roll could be used to deliver a whole replacement meter. This type of service granularity is common in IT and will probably have to be part of the service strategy for Utilities.

Oops - hit "enter" before I finished editing. Re Dell - the componentry of the model changes almost daily depending on what was available.

Final Comment: Applications drive hardware choices. Buying hardware today for applications that are not fully developed is a major no-no in IT. Applications are not developed on the oldest platforms in use - software developers are always keen to use the most powerful equipment available for development so as not to be unduly contrained. Just a no one is writing applications for the original version of Windows 2003 (less than ten years old), none of the software driving smart grid is going to be designed for platforms of that era.

Thus I don't think future proofing is realistic. Sorry.

Gay Gordon-Byrne,

I completely agree with you that "future proofing" smart meters for electric utilities is impractical and unrealistic.

In the consumer product world individual consumers pay the up-front costs for whatever level of future proofing they choose to pay for i.e. when they buy a personal computer system with hardware and software features a la carte. In the utility service meter world it is completely different situation because every residential customer typically gets equipped with the same meter configuration. So the utility must either bear huge up-front costs to equip all residential meters with all possible future-proofing options, or alternatively face replacing extremely large numbers of meters down the road. The latter is a bitter pill to swallow when they must pay for large-scale meter replacements with regulator-approved rate increases or extra charges added to everyone's bill.

This market problem could be reduced by changing the regulatory regime utility companies are forced to work under. Under the current regimes it must be either all customers getting a meter upgrade or no customers, there's nothing in their funding mechanisms to upgrade individual customers one at a time. Contrast this with our less-regulated CATV and telephone companies where these people have no issues with rolling a truck to a single customer willing to pay for an equipment upgrade in their home by a field technician.