The Smart Grid
Part One: The Challenges of a Modern Grid
On May 13th, 2009, the Utah Public Service Commission hosted a one day workshop on the smart grid. Most of the attendees were representatives from the industry. The non-industry attendees included a representative of Utah Clean Energy, an engineer from ATK Launch Systems representing an innovative approach to commercial energy use, manufacture, and grid integration, an economics professor from the University of Utah, and me.
The smart grid is an amorphous concept about a robust, modern approach to the manufacture, storage, distribution, and metering of electricity. There is no standard definition of the smart grid. At the moment there are around thirty federal, state, and municipal groups working on smart grid projects, and with about four and a half billion stimulus dollars available for smart grid projects, there are more on the way this very moment, I’m sure, so I don’t expect a completely smooth transition, or a clear, comprehensive set of standards by the end of this year when NIST ( National Institute of Standards & Technology ) is expected to have its proposal ready for review by FERC ( Federal Energy Regulatory Commission ).
The Energy Independence and Security Act of 2007 is, in its own words, an act “[t]o move the United States toward greater energy independence and security, to increase the production of clean renewable fuels, to protect consumers, to increase the efficiency of products, buildings, and vehicles, to promote research on and deploy greenhouse gas capture and storage options, and to improve the energy performance of the Federal Government, and for other purposes.” Within this broad scope is the mandate for smart grid development. The main premise of the smart grid is to avoid the need to add generating infrastructure—there are enough existing opportunities for efficiencies that new generating infrastructure is unnecessary and wasteful.
From the industry standpoint, there are fundamental challenges to be faced, the meter infrastructure, the software systems, integrating distributed generation, increasing renewable sources, load balancing and demand response under the new paradigm, reducing manufacturing costs, and dynamic rate structures that will be marketable to consumers. And these were what the workshop focused on, for the most part.
Metering received the most focus, which makes sense. Automatic meter reading ( AMR ) was a recent innovation of utility companies to make meter reading easier and less costly. This technology might have a place in the new infrastructure but for areas where it hasn’t been implemented, it will be pointless. Advanced metering infrastructure ( AMI ) will be used instead. Mention was made of “smart meters”, which would basically be the same as AMI but with more features. The basic distinction between AMR and AMI is that AMR is one-way communication while AMI and smart meters would enable two-way communication, as well as hour-by-hour or minute-by-minute usage tracking. The hope for AMI is that it might provide new capabilities like time-varying pricing options, usage data useful to customers and service providers, improved outage detection and response, and identifying and replacing equipment inappropriately sized for the load it serves. A FERC survey conducted in the first half of 2008 notes that 4.7% of meters in the US are “advanced” with the highest penetration in Pennsylvania ( 23.9% ), Idaho ( 13.8% ), Arkansas ( 11.3% ), North Dakota ( 8.9% ), and South Dakota ( 8.7% ). Consumer-owned utilities ( COUs or co-ops ) account for half of the advanced meters. 26% of the co-ops already have their outage management systems integrated with their AMI systems. Utah is one of two states in this survey ( the other is Mississippi ) that had no advanced meters.
Startup costs for software were initially dismissed as inconsequential until another presenter brought up one case where software implementation represented nearly 40% of total startup costs. The complexity of the myriad data required for the new systems makes this a significant part of any new infrastructure. Software costs over the long term might be of little consequence but initial outlays could make cost justification difficult in the beginning.
Security was mentioned, of course, but it wasn’t really explored beyond the security concerns of open source software and a very brief mention of microgrids. The gentleman from ATK Launch Systems provided details of their DOE funded project as an example of a microgrid. It includes the classic example of pumping the water up the hill during off-peak hours then letting the water flow down hill to generate electricity at peak load hours ( which is more about arbitrage and storage than security ). They also utilized a few other alternative energy sources like wind, which was abundant on their site. These systems provide up to 15% of their energy needs. This was mainly to cut their energy costs and carbon footprint, and it certainly does a good job of it, as well as providing reliable backup for a limited set of mission critical systems in case of outage, but I don't really relate to something like this as a microgrid since it’s just for one site. ( He also provided interesting insight into the problems they’re having with Rocky Mountain Power, which he characterized as reluctant and evasive. This was, of course, when the rep from RMP wasn’t around. )
It seemed that the industry was using reliability as a proxy for security and focused on this a bit more. But the main point here was that there is no standard evaluation of power quality. Reliability is more than just avoiding outages, or recovering from them quickly, it's also about the general consistency of the power delivery. Here, the presentation verged on highly technical discussions of waveforms and PMUs, and I hadn't had my cookie yet, so it’s all rather fuzzy for me.
Another issue that was only briefly mentioned was enabling new markets, for the utility companies themselves as well as third parties. Google was a ready example but I’m actually more concerned with front end markets, like Gridpoint. Gridpoint appears to be positioning themselves as a service provider to the utility industry by providing an integral product, the Gridpoint Energy Manager, along with customer services ( acquisition, support, and maintenance ), and a full range of services for the utility companies with economic analysis and financing ( through Goldman Sachs Group ), project management, integration, and installation; basically leaving the utility to simply manufacture power and maintain their part of the infrastructure.
Renewables were mentioned, mostly focused on the cost challenges of integrating more renewable sources. I was surprised to learn that there are diminishing returns for such integration. At about 30% to 40% renewables, utility companies will hit that wall. One presenter mentioned that the issue can be addressed, just that it will be difficult.
Distributed generation on a large scale envisioned for the smart grid is probably a little too far away to have been taken too seriously, so it was pretty much glossed over. Aside from diminishing returns, there are basic physics and logistical problems that need to be addressed with multitudes of small scale generator nodes feeding back into the grid intermittently. But this brings me back to microgrids. I see decentralization through microgrids as more fundamental to security than network software security. And microgrids will depend on distributed generation along with the ability to dynamically connect and disconnect to the grid at large. Since the distributed aspects of the smart grid are most related to planning issues, I would have preferred a lot more attention on microgrids.
Very little was said about distributed storage, either. Hybrid electric vehicles were mentioned passingly but nothing was said about lithium ion batteries, nickel-metal hydride batteries, SMES, flywheels, or supercapacitors. So I imagine nothing much has changed on this front.
Consumers don’t get off easy in all of this, either. The complexity of choices will create a barrier to adoption as most people might simply choose to stick with a flat rate on usage. The adoption benefits of an opt-out system were discussed but they didn’t really explore the possibility of consumers who aren’t interested in maximizing the benefits of variable rates eventually finding that standard rates would save them money with the effort to opt-out significantly less than managing the new technology. They readily and frequently admitted that self-selected pilot programs are not representative of potential consumer behavior overall. The rep from Rocky Mountain Power did mention their concern for the opposite disposition, though. She mentioned the possibility of reduced revenue due to increased consumer conservation. The only implication I can get from such a concern is that they are not interested in creating any new services to recoup this lost revenue. This wasn’t representative of the workshop, though. All the other industry reps seemed to take it as a given that consumer conservation is inevitable, and desirable, and new services would need to be created.
Another significant challenge for consumers is equipment replacement costs. Smart home appliances for a Home Area Network setup will need generous incentives for pretty much everyone. Smart thermostats can be relatively inexpensive and many already exist today. The problem is with everything else. You will need to replace every appliance in your home to get the maximum benefit. Apartment complexes will need to replace hundreds of every standard appliance. ( Knowing landlords, they will probably charge a premium for apartments so equipped, effectively offsetting any potential savings in energy costs. ) And since energy isn’t the only industry looking to realize the efficiencies of applied modern technology and systems design, consumers will have concurrent pressures to upgrade the other systems in their home. On top of all that, there's also the issue of interoperability. Here's the list of HAN protocols noted in the presentation: Bluetooth, FlexNet, HomePlug, LonWorks, 6LoWPAN, WiFi, ZigBee, ZWave, RDS, and Insteon. None of this means it can’t happen, it just means that adoption inevitably will be slow.
Home Area Networks also present potential privacy issues. How much control over your appliances would you cede to your utility provider? Would it be okay with you if they could turn off your air conditioner at peak load, basically when you want it the most ( which is why the peak load is created in the first place )? They’ll give you a discount rate for this lack of control, certainly, but would you take it? How much data would you be willing to let your utility company, or worse, some third party service provider, collect from your smart devices?
Remote activation and shutoff presents potential problems for individual consumers but the main problem it presents affects all consumers and the industry itself. The justification is that turning on services would cost the utility company much less with remote activation. ( I’m sure they’d still charge you a convenience fee. ) The problem is with security, though. Remote activation/shutoff will necessarily be a network software solution. I shouldn’t have to say anything more on the potential vulnerabilities this presents. So I won’t. Remote shutoff also presents the problem of utility abuse. This concern is usually presented by consumer advocacy groups.
The privacy issues from HANs and the fairness concerns of remote shutoff should be easy to overcome. The security concerns of remote shutoff will be more difficult to address.
Consumer participation could have other unintended consequences that may seem to be counter-productive. If given the chance to purchase power through the cheapest providers according to certain automation systems or utility-provided rate structure, consumers might choose this often enough as to significantly increase coal use, since it is the cheapest way to make electricity. Now, the smart grid isn’t necessarily concerned with environmental impact or being “green,” per se—it’s more about exacting efficiencies from existing infrastructure and providing better quality power—so it may not be so concerned about an increase in coal use, but the rest of us should be. And the best way to ensure the industry does care is to monetize carbon. To be fair, this actually is a concern for the research end of the industry at least, though one of the reasons is because they see monetizing carbon as very likely. Roger Levy from the Smart Grid Technical Advisory Services Project at Lawrence Berkeley National Laboratories gave a presentation that was particularly focused on the carbon savings to be had from a smart grid.
And having said that, I’m not necessarily advocating for carbon credits and a cap and trade system. As much as I want to be, I’m still not convinced. Something definitely needs to be done to create some kind of penalty for carbon. There are enough savings to be had simply from maximizing resource usage that this should be a no-brainer business decision, but all this reveals is the competence of our basic corporate mindset. Individuals seem no better, either—okay, maybe slightly better.
The second half of the workshop was in two parts: a survey of existing smart grid pilots and projects in the US and the EU, and a regulatory road map for Utah’s smart grid implementation. Unfortunately, I think I broke my brain writing this post, so I’m going to take on the second half of the workshop in Part Two: A Scramble for Standards & Stimulus Money. It should be much shorter.