Last update: May 15, 2013
How will consumption change? How will generation change? What unexpected efficiencies will be gained? And what does this have to do with renewables?
Smart grid technologies can change peak demand deferring major capital expenditures, reduce enormously the manual labour required to manage the grid and prevent many more expensive grid failures. They aren’t needed for low penetrations of renewables, but a subset provide advantages with higher penetrations.
Image courtesy: http://deepresource.wordpress.com/2012/04/03/smart-grids/
Smart Grid is a catch-all phrase for a bunch of technologies spanning generation, transmission, management, distribution and consumption. That’s a huge range.
There are three general functions to a smart grid:
- Instrumentation – This involves digital sensors connected to central management or each other being placed on many more technical components of the grid. Data from them is much more frequently brought together to allow:
- Analysis – Some analysis is very simple threshold tests for voltage, on / off state or dangerous situations. Other analysis is sophisticated calculations on the major data sets to determine long-term trends and hidden issues. This leads to the ability to do the right thing with:
- Control – The final step in a smart grid is acting on the information. Distributed and centralized intelligence, autonomous and under direct human control, adjusts the functioning of generation, transmission, distribution and even consumption through the digital network. Voltage spikes are managed down without human intervention. Consumer appliances such as air conditioning may be shut off. Natural gas generators are spun up for peak power needs and shut down when unneeded. And all of this would occur mostly without humans having to flip all of the switches and turn things on and off manually.
There are a few ways that smarter grids add value:
- Greater resilience means fewer failures means lower cost
Toronto 2003 blackout courtesy: http://en.wikipedia.org/wiki/File:Toronto_ON_2003_Blackout.jpgRisk and impact:Whether a transmission line goes down from a nuclear reactor or major supplying neighboring geography, or a major capital power plant goes dark unexpectedly, or a storm blows trees down on a major distribution point, the energy grid can see significant surges and dips in power. If unchecked and unmanaged, these dips and surges can cascade through the system, blowing circuit breakers and shorting out various important bits. This can lead, fairly alarmingly, to the eastern seaboard of the USA going dark, putting perhaps 100 million people at risk of freezing in the dark.Smart Grid Value:A lot of smart grid technologies are just monitoring equipment spread all over the place feeding data into distributed and central management areas, allowing rapid responses to local failures that prevent wide-scale grid failures. These responses can include bringing more peak generation online, buying a lot of power from a neighboring jurisdiction, isolating a blackout area from other parts of the grid, or routing electricity along a redundant path.Efficiencies:
As energy is made up not just of the cost of generation but also of all of the other components and maintenance of them, and staff overtime for line repair and the cost of regulation when something goes wrong, avoiding major failures has signficant efficiency value. When grid managers are doing their jobs right, they aren’t spending their time justifying themselves excessively to regulators. When widespread failures are avoided, massive overtime costs and likely medical claims due to tired crews having accidents are avoided. When expensive capital equipment is shielded from rapid shutdowns or startups, expensive maintenance and readiness / safety inspections are avoided. As an example, if a nuclear reactor has an emergency shutdown, something called a poisoned core is a typical outcome; this can lead to the reactor being out of service for six months or a year easily, with millions of dollars of service costs.
- Greater electronic, remote monitoring means less manual monitoring therefore lower costs
Manual version:WIthout smart grid technologies, you have to employ a lot of people to look at dials and meters spread all over the power authority area. As an example, manual meter reading in Quebec, Canada happens six times a year and has a cost of about $6 per meter read for a total cost of $350 million per year. Meter readers are unionized and well-paid to drive around, look at little dials and write down numbers. This is one example of all of the monitoring typically done by humans of the enormous swath of generation units, transmission lines, transformers, distribution points, distribution lines and other grid assets such as the fleets of vehicles necessary for all of these people driving around.
Smart Grid Value:
Putting smart meters on each house enables meter reading every 15 minutes and no visits by manual meter readers. An entire fleet of vehicles taken off of the road. An entire workforce freed to do something more productive for society and likely themselves. A significant health and safety risk due to accidents and dogs reduced, increasing overall productivity.
There are some ancillary costs, but still Hydro Quebec has the potential to save $300 million in labour and fleet costs by putting in smart metering. That’s without saving any money by people changing their habits (coming up).
- Time-of-day pricing leading to changes in consumer behaviour leading to levelling of peak consumption and deferred capital generation projects
Current state:As manual meter reading typically occurs monthly or bi-monthly, there is no ability to charge consumers anything except a flat rate for electricity regardless of when they use it during the day.
Smart Grid Value:Smart metering enables consumers to know what they are consuming in much more granular ways either through dashboard appliances in their homes or through portals (much more common and valuable as a solution now). Time-of-day pricing charges differentiated rates to consumers: more at peak times, less at non-peak times, lowest at trough times such as the middle of the night.
A small percentage of the populace will, without anything other than that knowledge, change their behaviour to reduce their overall consumption. Add time-of-day pricing and make that visible on people’s bills and all of a sudden larger changes in behaviour start to occur, especially in high-electricity cost jurisdictions. People put use timers to bake stuff, run dishwashers or do laundry in the middle of the night. People open windows for breezes rather than run their air conditioners full bore at 6 PM. People who do this in high-cost jurisdictions save a ton of money on their electricity bills and tell their friends, neighbours and co-workers, who also do it. This is great for them as consumers: lower costs. But it’s also great for the grid management people. All of a sudden, a bunch of the peak demand has shifted to non-peak hours. That is really, really valuable because the grid management folks have to have sufficient generation capacity for the peak plus emergency backup; shifting demand out of peak means that they don’t have to build more generation capacity, or can shut down the least efficient generation capacity.
- Better and more frequent electronic monitoring makes asset maintenance more efficient and effectiveCurrent State:All of the assets of the grid — generation plants, lines, transformers, etc — are manually inspected on a sub-optimal timetable; the grid management people balance frequency and quality of data with cost of acquiring the data. As a result, they are dealing with sub-optimal information that doesn’t allow maximal optimization of asset maintenance, which can be really expensive. As a result, they maintain many assets more than is strictly necessary, at great systemic expense. And some they under-maintain through lack of insight into edge wear conditions leading to expensive failures.
Image courtesy of http://www.smartgridnews.com
Smart Grid Value: Instrumentation of assets and electronic capturing and centralization of assets provides up-to-date information and much more of it on assets. This allows much more systemic analysis of that information. This allows much more optimization of asset maintenance, downtime and much better prevention of failures.
Assets last longer, operate for more of their lifespans and fail much less often, requiring less expensive emergency repairs. Tons of money here.
- Grid management automation reduces the need for manual grid management staffCurrent State:Somewhere in every jurisdiction there is a building or several buildings full of very bright, very educated, expensive people. Many of them are approaching retirement, or working on contract after retirement. They gather all of the data related to the grid. They author reports. They think of strategies to improve the grid. They respond to emergencies. And every day a whole bunch of them are responding to fluctuations in demand, generation and grid stability in real-time; three shifts worth of them in fact. With backup. And failover. And vacation staffing plans. And then there are all of the people that they talk to in the field who manually adjust dials and start and stop generation plants manually.Smart Grid Value:Distributed intelligence in the grid means that current and next generation distribution points can manage many voltage fluctuations in their small portion of the grid without human intervention of any sort, damping oscillations and keeping electricity humming along at the right pitch. Centralized grid management intelligence systems can automatically deal with many grid destabilizing situations without human intervention, asking wind farms via SCADA-interfaces to feather their blades for example, or automatically drawing down on energy contracts from neighbouring jurisdictions, or adjusting the settings on distribution points to ensure that energy stays clean. They can produce all of the reports and data analyses that the smaller number of people need. And they can alert the smaller number of people much more rapidly when something that truly requires human input and intelligence is required.
Image courtesy of http://www.sustainablebusinessoregon.comEfficiencies:
Lots fewer expensive people sitting around doing stuff that computers can do both much more cheaply and much better. Lots fewer staff driving out to distribution points in fleet vehicles. Lots fewer fleet vehicles.
- Renewables Integration;Renewable energy sources vary more and more predictably than nuclear or fossil fuel sources, whether it is seasonal water levels impacting hydro generation, night preventing solar generation or calms impacting wind generation. Changes in production of nuclear or fossil fuel generation are due to major transmission losses such as the Hunterston Nuclear Plant in Scotland or damage to plants requiring extended repairs such as with Bruce Nuclear. Those predictable but regular variances could destabilize a grid if there were sufficient renewables on the grid and the grid wasn’t being managed carefully.Current State:Workers might have to be dispatched to adjust substation parameters. Wind farms might be called and asked to feather their wind farm or a portion of it, and the wind farm operator might dispatch workers to do that. Passive hydro — unused water capacity behind existing hydro dams — might be allowed to build up requiring adjustment of water volumes at dams in widely distributed areas. Natural gas peak supply generators would be spooled up; a worker would be required to turn the switch.Smart Grid Value:All of the above could be accomplished through SCADA control interfaces without manual intervention. This reduces travel time, staff overhead, staff safety impacts and additional human resources costs. And the reaction would be swifter and more accurate, ensuring greater grid stability with greater penetrations of renewables.
Beyond the obvious human resource cost reductions, increasing renewables penetration has significant advantages as coal, for example, as negative externalities that are priced at 17.8 cents per kWH for health and environmental impacts; natural gas might be in the 5-10 cents per kWH range. The more renewables on the grid, the less fossil fuels that will need to be burned and the better off the people supplied by the grid and the world will be.
It’s also worth noting that smart grids reduce human deaths, accidents and other situations requiring medical treatment and hospitalization. They do this directly by reducing the number of times workers have to do routine maintenance and by reducing the number of times workers have to do emergency fixes; both of these types of work lead to workplace accidents and when you are dealing with potentially hundreds of thousands of volts and 100 meter heights, minor accidents kill. Smart grids do this indirectly by reducing power outages and allowing people to have cooling on lethally hot days and heating on lethally cold days at a lower price and with greater certainty. They ensure that people have lights when they need them so they have fewer accidents in crisis situations.
Smart Grids are already here
Finally, grids have been becoming smarter for decades, mostly starting in the major generation and transmission side of things. SCADA interfaces have been around for a long time. Instrumented reading of remote systems has existed for a long time. All of the smart people I referenced above have been exploiting efficiencies to bring better value to their stakeholders and consumers for a long time. Smart grid is just them working with major vendors such as GE and IBM to continue doing what they have been doing: exploiting advances in technology to make the grid better for everyone.
- For an excellent series of videos on smart grids, please see the DeepResource blog post on them here: http://deepresource.wordpress.com/2012/04/03/smart-grids/