Tuesday, July 12, 2011

California Utilities Report Smart Grid Metrics




The smart grid has attracted a lot of public8ity as it is a good idea that is actually feasible today.  The catch of course is that with all such processor driven applications, the hardware rollout is the trivial part and it gives everyone a false sense of accomplishment.

The hard part is the software rollout is way more difficult and it is normally an evolutionary process.  It goes from crummy (today as in this material) to a good working system in months and actually years.  Recall that it took most of twenty years for complaints regarding operating systems to largely subside.

The smart grid promises to be just as demanding for programmers as hardware will now respond to the new capability by changing.

Again the idea is good and it is timely.  Just do not expect too much too soon.

California Utilities Report Smart Grid Metrics

KATHERINE TWEED: JULY 8, 2011


Buried amongst 900 pages is information about how California’s Big Three are delivering smart grid value to their customers.

When California’s three largest utilities were ordered to file smart grid deployment plans, it was expected that some sections would look very similar since San Diego Gas & Electric, Pacific Gas & Electric and Southern California Edison worked together while creating the filings.

From distribution automation to electric vehicle adoption, there are similar threads throughout the reams of paper filed this month with the California Public Utility Commission. There is one section, however, that all three utilities agreed on fully: metrics.

California is unique by not just calling for utilities to lay out their grand smart grid plans but also asking them to quantify and measure progress. As advanced as California is with its smart grid ambitions, the initial reported metrics show that there is still a very long way to go in reaping the benefits of the investment, especially in the customer realm.
The metrics, which were developed over the last year, will be used by the PUC to provide updates to the legislature. Even though they won’t be finalized by the PUC until later this month, there will likely be little change from the current iteration. The metrics cover customer and meter metrics, storage, electric vehicle integration and reliability. SCE gets style points for the most color and providing ample explanation along with the figures.
Although there are some interesting data points in the reliability and distribution automation metrics (SDG&E already has 73 percent of its circuits automated, compared to less than 50 percent of the other two utilities), the best area to watch will surely be how the three utilities deliver the value of the smart grid to their customers.

Here are the metrics to watch:

1. Number of advanced meter malfunctions where customer electric service is disrupted

SCE wins on round one of this measurement with a total of zero for 2010. The figure does not include meter malfunctions that result in wonky billing, just those that occur when service is interrupted. PG&E put up the largest figure with 587 meters. That was attributed to radio frequency interference with breakers, an issue that the utility says it resolved by installing low power meters. SDG&E reported 37 meter malfunctions with no explanation. It would help to have a baseline to compare amongst traditional meters to better understand what these numbers mean for consumers. However, a few hundred meters is a small figure given that millions have been deployed. 

2. Load impact from smart grid-enabled, utility administered DR programs

Demand response is already happening across California, but the initial figures of programs that have been enabled because of smart meters, rather than legacy interval meters (mostly on the commercial & industrial side), is limited. SCE could not clearly attribute any of its DR programs from large C&I programs directly to smart grid devices for 2010. PG&E was in the same boat as SCE for 2010, although it did not specify if its smart meter rollout would enable more DR in 2011. SDG&E counted 50 MW, 14 of which came from the residential sector.

3. Percentage of demand response enabled by AutoDR

California is the only state that has officially adopted OpenADR, or open automated demand response, as a standard. Interestingly, PG&E did not measure loads that were enabled by AutoDR versus other demand response platforms. SCE had the largest reported value, with nearly one quarter of its Demand Bidding Program coming from AutoDR. Ten percent of SCE’s critical peak pricing program came from AutoDR, compared to 4 percent for SDG&E. Those figures should continue to grow in 2011, especially since this is a separate metric from more general load shedding. Various companies, from EnerNOC to Honeywell, have embraced AutoDR, with the expectation that it will expand within and beyond California.

4. The number of utility-owned advanced meters with consumer devices with Home Area Network (HAN) or comparable consumer energy monitoring or measurement devices registered with the utility

To sum it up, the figure here is zero all around, with the exception of a few very small trials conducted by SCE and PG&E. The utilities are waiting for the adoption of the Smart Energy Profile 2.0 HAN national standard, which might not be finalized until 2014 at the earliest. SCE, however, says it will have a pilot with several thousand devices later this year. The other utilities did not attempt to explain future rollouts or setbacks, but it is likely that large-scale HAN hardware is still years in the future in California.

5. Number of customers that are on a time-variant or dynamic pricing tariff plan

One factor that might drive No. 4 is dynamic pricing. In 2010, the figures varied wildly when it comes to pricing. PG&E boasts more than 100,000 residential customers on a dynamic pricing tariff, although they don’t break down which of those are in a peak time rebate program versus a variable time-of-use rate plan. SCE counts zero, with programs beginning in 2011. SDG&E is somewhere in the middle with 1,888 residential customers on dynamic priding and nearly 25,000 C&I customers.

This metric is interesting and vague at the same time. There are various types of programs, from rebate programs to time-of-use rates, and as it stands, the specific breakdown is not reported. However, the right pricing schemes can help drive people and businesses into demand response programs, which in turn can lead to more adoption of HAN devices on the residential side -- so this is one metric to keep a close eye on.

6. Number of escalated customer complaints related to the accuracy, functioning, or installation of advanced meters

This is a murky metric because it starts with “escalated,” which does not include the initial complaints that come into a call center. In one way, it is more accurate because there are various complaints that are resolved in a single phone call -- but it might be telling to see the difference between initial and escalated calls. It is not PG&E that leads this category, but rather SDG&E, with 2,123 escalated calls. PG&E is second with 1,470.

As the utilities provide customers with access to usage data, these numbers should shrink in coming years. This is a metric that is already being used in a different way in Texas, which has a deregulated power market. On the Power to Choose website, where customers can choose their retail electricity provider, consumers can see how different companies rank in terms of monthly complaints (not just about smart meters). Although it is a little crude, giving consumers information about how easy it is to do business with a particular utility or company is a metric that many other deregulated states could benefit from.  

7. Number of utility-owned advanced meters replaced annually before the end of their expected useful life

This is another metric that, for 2010, looks like it was sliced a few different ways. SDG&E reported more than 27,000 advanced meters were swapped out in 2010, compared to just 4,739 for PG&E. However, PG&E noted there were another 50,000 advanced meters that were removed for investigation under the “meter vendor warranty process.” PG&E said 22,000 of those might have malfunctioned. SCE had about 8,500 advanced meters that it replaced. Although 20,000 meters is a small percentage of the millions being deployed, that’s still a lot of households -- and potentially angry individuals -- with a faulty meter.

8. Number of advanced meter field test performed at the request of customers pursuant to utility tariffs providing for such field tests

PG&E leads this category with more than 10,000 customer requests for their smart meters to be tested. The other two utilities each had fewer than 1,000. Last year saw multiple reports that validated the accuracy and safety of smart meters, so all of the figures should shrink in 2011.

9. Number and percentage of customers with advanced meters using a utility administered internet or web-based portal to access energy usage information or to enroll in utility energy information programs

PG&E also jumps ahead in this category, although the other two utilities are already have plans underway for portals for 2011. PG&E reported more than 400,000 usage inquiries for smart meter customers. All three utilities are offering portals for customers to check their energy consumption, and the offerings should become more sophisticated in coming years. The California PUC also recently laid the groundwork for third-party access to smart grid data, so other players will be looking to come into the market as soon as those are finalized.

For some reason, plug-in electric vehicle price tariffs garner their own category in the metrics section. PG&E is out ahead in this as well, with nearly 200 customers on one of its two EV tariffs, compared to about 120 for SCE and 30 for SDG&E. However, SCE notes that for those with plug-in hybrids, specific rates might not appeal to those drivers because of smaller battery sizes and lower charging levels. These figures don’t accurately reflect EV adoption in California, and the figures might not be reflective of a utility’s ability to handle EV load until years down the road.


Comment by Carl Hage 07/8/11 3:16 PM

I’m in PG&E territory, and I and my friends have new “smart” meters. But I can tell you that it is impractical for customers to get detailed usage information from the PG&E web site—it’s a horrid pile of complex javascript, flash programming, overly complex cross-site CGI scripts with thousands of bytes of cookies. (It must have been expensive to program.)

The claim is that smart-meter information is available, but it takes way more than 5 minutes of clicking to get a week’s worth of data—if you are lucky and the scripts don’t hang. (I’ve never been that lucky and haven’t had the patience and persistence in retrying to download just a weeks worth of data.) Even the most eco-conscious people cannot get at the data needed to analyze usage and get estimates of cost savings. Third party access is not allowed so Google power meter couldn’t be used for their own employees. People have to buy a $300 add-on because the PG&E web site sucks.

I’m working on a set of scripts to hack the PG&E web interface so I can download my own data and data on behalf of my friends, but it’s really complex and I haven’t had time. Another guy did the same (but in Ruby) and commented on the complexity. There is a problem with using the software on a third party site because you need your billing login and password to access meter data—I can only do this for friends that trust me.

Here are my suggestions:

1. The CPUC should mandate that customers should have convenient access to detailed usage information and summary reports. Customers should be able to use third party software or services without compromising billing passwords. Pricing information (for use by third party apps) needs to be in computer readable format (not piles of print-only PDF).

2. The CPUC should form a panel to evaluate the usability of web/internet access portals to usage information, and evaluate the effectiveness of summary reports and paper bill inserts. Utilities should be given a grade A-F: A means hard to imagine better, F means unusable. Utilities with an F must scrap the system and start over (and perhaps disallow cost recovery). (PG&E gets an “F” because it’s so complex and tedious no-one can use it to really analyze data, and the graphs don’t really help people identify costs and possible savings. The poor interface inhibits third party analysis.)

3. Customers should be able to create a separate login ID and password only used to retrieve usage, so it can be used with third party software and services to download data. Customers (using billing login only) can change this password to rescind access.

4. Customers should be able to sign up for a weekly or monthly email message to have usage data delivered directly in email attachments (no web access). Besides data viewable in a spreadsheet or auto-imported into a software tool, the message could include some summary analysis.

5. Utilities should develop sophisticated statistical analysis software to recognize certain usage patterns and guess at possible cost-saving opportunities. For example, my refrigerator has a square-wave regular usage pattern especially visible at night—software analysis of a month of data can estimate annual cost for running this refrigerator and compare it with new refrigerators. Usage by hour-of-day (weekday+weekend) can be summarized with 10%-50%-90% ranking. Looking at changes during evenings, etc. estimates could be made for possible savings with different lights, new energy-star TV, etc. Vampire power can be easily reported with annual cost. A friend has >200W of base power—way more than needed. Since utilities have meter data for all houses, it’s easy for them to compare a customer’s use patterns with a histogram of all houses in the neighborhood. The analysis could be improved if customers opt to go to a web site and give more information, e.g. how many people live in a house (adults/teens/children), what kind of appliances, heating, cooling, etc. are used, is there a pool, etc.

6. The paper bill should include a basic report on usage, e.g. small graph of daily usage, and a graph (low-median-high) showing usage by hour. A summary could estimate annual savings opportunities for reducing vampire power, changing lights, etc., perhaps as a percentage and compared to neighbors. The report could include if time of day billing is better or worse than flat-rate, or could be better if usage patterns were shifted. This simple report included with the paper bill could include an invitation for more detailed analysis. Customers with excess use patterns could have special messages. People with e-bills could receive even nicer reports as email attachments.

7. The standard analysis/reporting by the utilities included with paper/electronic bill should compete with third party analysis software and services. Data must be easily accessed by third parties and under control by the customer.

No comments: