Wasting electricity

The New York Times is running a series on data servers, the massive centers that house bits and bytes generated by millions of Internet users. They started out small a few decades ago; today, they are massive and proliferating.

Though I am intimately engaged in energy issues from my perch on a utility board, I was unaware that, according to the Times, these data centers waste more than 90 percent of the electricity they consume to perform their intended function. Worse, they rely on dirty diesel backup generators, which themselves are giant and plentiful. And we should use “backup” advisedly. The Times:

Energy efficiency varies widely from company to company. But at the request of The Times, the consulting firm McKinsey & Company analyzed energy use by data centers and found that, on average, they were using only 6 percent to 12 percent of the electricity powering their servers to perform computations. The rest was essentially used to keep servers idling and ready in case of a surge in activity that could slow or crash their operations.

Microsoft has built huge server farms in Quincy, Wash., to take advantage of cheap electrical power provided by Grant County Public Utility District, which owns two of its own dams on the Columbia River that generate nearly 8.5 million megawatt-hours a year. The town is small, with fewer than 7,000 residents. It’s a farming community, both traditionally and technologically.

Before the arrival of Microsoft and Dell and other Internet behemoths the town’s residents and agricultural enterprises consumed about 9.5 MWa, or about 83,000 MWH. Last year Microsoft and Yahoo combined to consume 41.8 MWa, or roughly 366,000 MWH.

If the Times‘ article is correct on the waste numbers, and if Microsoft et al. continue adding capacity, then there could be implications for the region down the road, especially if the economy recovers and demand for all things increases. At the moment, we may be collectively unmoved, because there is so much supply of electricity amidst stagnant consumption. But if demand rises, those waste figures could loom large.

Right now, the electricity is cheap. The Times‘ article says that the technology giants pay between 2.5 and 3.8 cents per kilowatt-hour. Yet, even at those low rates, there’s got to be opportunity for efficiency measures that are much less expensive than the power.

I predict, then, that sooner or later a niche of energy conservation firms will step in to sharply diminish wasteful consumption and reliance on dirty diesel backup. The region, certainly, would be better off.

Tax the things you loathe

America’s love affair with all things fossil comes with inconvenient consequences. I’ll list a few:

  • carbon dioxide emissions
  • dirty air and water
  • cars
  • roads
  • sprawl

One could add to this list obesity and other health assaults, since spending so much time behind the wheel (rather than walking or bicycling) contributes to expanded waistlines and lousy circulatory systems. Add in the mobile Mickey D with fries and you’ve laid out a path to an early grave.

Economists, for the most part, recommend that we tax carbon. But Americans, for the most part, hate taxes, despite our paying fewer than our European counterparts. So, whenever a well-intentioned politician (small in number, certainly) proposes a carbon tax, he or she invites failure; colleagues won’t touch it.

However, we do love our subsidies and expect our own politicians to deliver them in support of local industries. Should we live in coal states like Virginia we appreciate whatever help the federal government can provide, whether in tax relief, subsidies, or fewer environmental regulations.

So, let’s take a look at energy subsidies over time (source).


Oil, natural gas, and coal receive the most help from Congress, with oil the runaway winner. It should not surprise us, then, that we use more of these resources than we would without the subsidies. Similarly, if subsidies increased for the green resources, we’d expect greater development of renewables.*

It makes economic and environmental sense to both eliminate all energy subsidies and impose a carbon tax. From the sourced article:

We can either tax carbon or incentivize non-carbon. Debate rages about which is better, but it has always been more effective to tax what you don’t want, rather than to pretend to know what will work in the future.

If Congress imposed an effective carbon tax, we’d see much more innovation in alternative fuels (e.g., geothermal and solar) and greater investments in demand-side resources, including conservation and distributed generation. But I said “effective” tax. It must be high enough to approximate the (negative) value of carbon’s pernicious consequences. (I’ve written about his here.)

I am not hopeful of seeing either a carbon tax or some cap-and-trade scheme in my future, which gets shorter by the year. Fossil-fuel interests are huge and politically well-connected. Moreover, we don’t like taxes. We’re suspicious of government. We’re dangerously ignorant. This unholy combination yields a toxic business-as-usual strategy that promises profits for the few and catastrophe for the many.

Got to go. Time for my walk.


* I am forced to quibble with the “market activity” subsidy for hydro. That’s a euphemism for a controversial, yet longstanding function of federal marketing agencies like BPA. By law, and this goes back to the New Deal, the output from federal systems (in our case, the nearly two-dozen dams along the Columbia and Snake rivers) must go first to public utilities (e.g., Snohomish County PUD) at cost (technically, enough to recover Bonneville’s financial obligations—its revenue requirements). For years, politicians not served by the FMAs have demanded that the FMAs charge market-based rates. Not only would such a policy-shift (requiring a change in the law, by the way) jump up electricity rates in Washington and other states receiving preference power from the FMAs, it would generate more revenues for the federal government—a wealth transfer representing billions of dollars. We should keep in mind that Bonneville’s customers have fully funded the FMA’s statutory obligations over seven decades. To those who might argue that having to pay market-based rates (higher the current preference rates) would spur more conservation, please understand that the Pacific Northwest, along with California, leads the nation in conservation acquisitions. The Snohomish County PUD, for one, has set annual records over the last several years.

Lost and gained

The Snohomish County PUD was fortunate enough to land Bob Bolerjack for its corporate communications department. He had previously been the editorial page editor for the Everett Herald. I thought Bob was as good as they come writing for the newspaper.

He transferred from the sports section to the editorial pages just before I made my first run for the PUD board, back in 2002. I couldn’t have written a better endorsement of me than Bob’s. In his years as editor he remained intensely curious about energy matters in general and the PUD in particular, attending many utility events, including several board meetings, to learn more about the PUD’s conservation and renewable energy initiatives. He was impressed enough to give me a second glowing endorsement in my re-election campaign in 2008, then just recently going all in by joining the utility.

Yesterday the Herald announced his replacement. Peter Jackson, son of the late Senator Henry M. Jackson, will assume Bob’s pen. The paper will hardly miss a beat, as Peter is also an accomplished writer, most recently for Crosscut, the online paper founded and published by former Seattle Weekly creator David Brewster. Peter is an avid outdoorsman and environmentalist. Years ago he championed the PUD’s conservation efforts, which have been rejuvenated under this board’s leadership and the utility’s ultra-capable staff.

I look forward to reading Peter’s insights. I’m sure that he will not disappoint.

Sunset Falls [u][u2][u3]

[u3] I was trying to take pictures while listening to all the facts and figures. It seems that I got a few items wrong, as AGM Kim Moore discovered. In the interest of accuracy, here goes:

  • on fish handling 10,719 was the record one day haul of pink salmon (9/14/2009)
  • entrance to ladder rather than intake to ladder
  • transform Sunset Falls AREA into a better place than it is now [davald: the falls themselves will not be altered]
  • the trap and haul was built in 1958 by Department of Fisheries and not Puget Power and Light; the land was previous owned by Puget Power and Light and intended as a hydroelectric project
  • more than one home can see the intake [davald: a few will be able to glimpse a side view of the intake]

[u2]Video at the end.

[u]Correction below.

The Snohomish County PUD has a rather simple resource strategy: capture all cost-effective conservation and, if necessary to meet increasing load, develop renewable resources in our own backyard. This strategy allows us to comply with Initiative 937, reduce financial risk, add resources in small increments to match load growth, impose a smaller ecological footprint, and keep costs, and therefore rates, stable.

Over the last five years the PUD has broken its own historical records for annual conservation savings. Last year the utility saved 10.1 average megawatts.

On the resource front we’ve been exploring tidal power in Puget Sound, installing solar voltaic panels on rooftops, constructing new low-impact hydro projects (we’ve got mountains and rain in Snohomish County: a good combination for letting water and gravity generate electricity), and engaging in serious efforts to add a geothermal project to our portfolio.

As always, the bulk of our power—over 80 percent—comes from the Bonneville Power Administration. With the exception of the output from the state’s lone nuclear power plant (Columbia Generating Station), all of that electricity is generated by a series of hydroelectric projects along the Columbia and Snake rivers, including Grand Coulee.

We’ve begun exploring another hydro project, this one at Sunset Falls. It appears to be an ideal site. But by merely announcing our intentions to investigate we’ve encountered opposition from some of the local residents and various non-governmental groups. Unfortunately, much of their rhetoric is inaccurate, often wildly so.

This morning I joined my commission colleagues, the general manager, the clerk of the board, and the PUD’s assistant general manager in charge of generation in a visit of Sunset Falls and surrounding environs. We commissioners wanted to see for ourselves what the fuss might be about.

The falls are certainly spectacular. Here’s one shot I took from the shoreline opposite the falls.

Near the center of the photo is an existing intake structure and fish ladder. Puget Sound Power & Light The state constructed both about 50 years ago. The investor-owned utility, now known as Puget Sound Energy, did not build a hydroelectric facility. The Department of Fisheries uses the structure to capture adult salmon swimming upstream. The fish cannot negotiate the falls; they must be trucked up stream, where they spawn. The juveniles then make their way downstream and over the falls then eventually out to the Pacific Ocean.

Here is a closeup of the capture tank. Fisheries workers use nets to pull the fish out of the tank then place them onto a single, rather ancient truck. I was told that as many as a thousand adult salmon a day may be transported. Should we proceed with the project, we’ll greatly improve this structure to enhance both the aesthetics of the site and the capturing and transporting of the salmon.

Some of the opponents of the project, whose details are far from being finalized, imagine that the utility will construct a giant concrete dam to impound the water for hydrogenation. Hardly. The plans thus far involve constructing an inflatable weir across the stream, reducing flows by approximately 2,500 cubic feet per second. (Flows can exceed 80,000 during extreme flooding. This morning, I was told, the river was flowing at 6,500 cfs.)

Here’s a slide from an October 2011 presentation to the board. It’s a plan view of preliminary designs for the intake. The inflatable weir is the balloon-like object near the middle; the intake is at the bottom left.

I’ve been informed that only one local resident would be able to see the intake structure and the weir, which would be inflated during low-water periods. During the winter when the PUD would need the electricity, the weir would likely be deflated, since the water flows will be at their highest. (The weir would inflate and deflate automatically according to river flows. It will be monitored remotely in real time and can be operated via computer if necessary.)

Here’s a preliminary layout of the proposed project.

You’ll notice that the water flows around a peninsula. At the left-most point of that peninsula the PUD would construct the inflatable weir that would connect with the side opposite, where the intake structure would be located. The diverted water would flow downhill through a constructed tunnel under ground to the site of the current fish trap structure. There the PUD would build a powerhouse that contains the turbines for generating electricity.

Here’s a picture of the intake area from a location just to the right and near the same shoreline.

To the left of the picture is the tip of the peninsula. The water flows around the peninsula from the picture’s left to right, then toward the camera. It continues to Canyon Falls, which is above Sunset Falls.

As I said at the outset, the falls are magnificent. The PUD’s proposed project will have no effect on the appearance or operation of the falls, since the utility would divert only a fraction of the river flow above. No resident, with the possible exception of one, would see any difference after the project’s completion. That one resident, as I mentioned above, would have a view of the intake structure.

Construction, itself, is a different story, of course. Building things makes noise. There will be trucks and crews on site for a number of months.

But once the project is finished, the entire site will be visually improved. The power house, for example, will be built so as to resemble the existing rock formation the power house will replace. The Tulalip Tribes and government fishery agencies support the PUD’s plan to greatly improve the trap-and-haul component of the proposed project.

Nevertheless, the PUD, as always, will be especially sensitive to the concerns of those who now oppose the project. We’re still in the preliminary phases, gathering and studying data, meeting with local residents and other stakeholders. We have a well-deserved reputation for negotiating in good faith, accommodating the wishes of others. I can make this guarantee: should the project proceed, the PUD will transform Sunset Falls into a better place than it is now.

Renewables have not displaced fossil-fuel generation

Washington is different, the Snohomish County PUD even more so—in this respect: the majority of the electricity is generated by renewable resources, mostly hydroelectric plants along the Columbia River. There is only one coal plant still operating (Centralia), and that will be phased out over the next decade and a half. Natural gas facilities contribute a bit over ten percent (pdf) of electricity generation. With the passage of Initiative 937 (pdf), both private and public utilities with more than 25,000 customers (the PUD has over 300,000) must populate their supply portfolios with increasing percentages of renewable resources, save for new hydro, which does not qualify, for dubious reasons. The resource of choice is decidedly wind power.

Utilities are scrambling to meet the initiative deadlines. This year the percentage of the supply portfolio that must be in the form of qualifying renewables is three percent; it jumps to nine percent by 2016; and reaches its maximum of 15 percent by 2020. The PUD is well ahead of the curve.

Things are different for the U.S. as a whole and for over 100 other countries. A study by University of Oregon professor Richard York concludes that almost all of the renewable resources built over the last 30 years have been swallowed up by increasing demand for electricity; the green resources have not displaced coal and natural gas.

York found that each kilowatt-hour (kWh) of electricity generated from non-fossil-fuel sources displaced only 0.089 kWh of that from fossil fuels. To put this in perspective, displacing one kWh of fossil-fuel electricity would require over 11 kWh of electricity from alternative sources.

It’s not easy being green.


Light bulbs and taking the long view

By default I am enmeshed in energy issues, given that I sit on the board of the Snohomish County Public Utility District, which purveys electricity and some water to over 300,000 customers in Snohomish County and Camano Island. During my initial campaign for office in 2002 I pushed what I called ‘The Goal.’ In fine, the PUD would continue to meet the electricity demands of its customers with no net increase in aggregate load. Since the utility is adding customers each year, to achieve The Goal per-customer consumption would necessarily have to fall. That could occur only through steadily increasing energy efficiency.

Well, after being in office for nine years and a quarter, the utility has fallen short of meeting that challenge. Not for lack of trying, mind you, as each year the PUD acquires record amounts of conservation. But instead of declining per-capita consumption, we are seeing just the opposite. Here’s a graph showing annual per-customer consumption in the residential sector.

Moreover, total residential consumption (aggregate load) is moving further away from The Goal. The utility is connecting new customers and the average customer is using more electricity than before. The combination manifests itself in rising aggregate demand.

But I have not given up hope. I believe that The Goal is still possible. However, to do so will require a new paradigm in approaching conservation. I suspect that it will have something to do with taking a long view. I’ll focus on one example, the light bulb.

Humans tend to be an impatient lot. We would prefer to have something today rather than defer gratification. I’ve got money in my pocket. Where shall I spend it? Business owners, in particular, focus on the short term. They want a rapid return on their investment. So they’re typically unwilling to take the long view.

The PUD, on the other hand, can afford to take the long view. We’ve been operating for more than half a century; I expect that it will be around for many decades to come. Moreover, the PUD is a public, non-profit institution with access to tax-exempt financing. We should not be constrained by short-term considerations that inflict our customers.

Let me give you an example of the difference. It’s about a light bulb. Not your ordinary incandescent, mind you. This bulb would set you back $50. Manufactured by Phillips in Wisconsin, the bulb received a Department of Energy prize for lighting innovation.

The Washington Post quickly denounced both the award and the bulb, the latter deemed “too costly.” The newspaper included a graphic that purported to demonstrate that buying incandescent bulbs would be cheaper, even over the long run. The super-efficient, 10-watt bulb, which will last 30 years, would cost the buyer $5 more than had he purchased a series of incandescents.

There is a big problem with the arithmetic, however. The Post got it wrong, decidedly so. The Post assumed an electricity rate of one cent per kilowatt-hour. Yet, the national average retail rate is about 12 cents/kWh. (The PUD rate is about 8.4 cents.)

A corrected version of the Washington Post lightbulb cost comparison shows $50 LED bulb over $100 cheaper than incandescents.

Instead of costing the consumer $5 more, the bulb would actually save $100.

Yet, who wants to spend $50 on a bulb, even if it will last three decades and save you $100? Heck, the bulb would outlive me. If most people had the same attitude, the Phillips bulbs remain on retail shelves. Enter the PUD.

Suppose there are 300,000 residential PUD customers and each has an average of 10 light bulbs. That’s a total of 3 million bulbs. Suppose further that the average wattage is 60 and that each bulb is on for five hours a day. (The assumptions are just for discussion purposes and don’t reflect actual data, which I don’t have in front of me.) We’ll also assume that the retail rate is eight cents per kilowatt-hour.

3,000,000 bulbs x 60 watts x 5 hours = 9 million watt-hours, or 900,000 kilowatt-hours
900,000 kWh x $0.08/kWh = $72,000 per day

Now let’s suppose that all these bulbs are replaced with super-efficent Phillips bulbs.

Three million bulbs @ 10 watts each running five hours a day = 150,000,000 watts, or 150,000 kWh/day. At eight cents/kWh, the daily total would drop to $12,000, for a savings of $60,000/day.

A 60-watt incandescent bulb lasts about 5,000 hours. The new Phillips bulb stops working after 25,000 hours, or five times as long. Thus, we’d need to buy five incandescent bulbs for every Phillips bulb.

Suppose the incandescent bulb costs 53¢. Right now the Phillips bulb would set you back $50. You’re not likely to shell out that kind of money for a light bulb. If you bought 10, you’d be out $500. For that you could purchase almost 1,000 incandescent bulbs. But bear with me.

If you purchased the Phillips bulb you would have saved $57.65 over its life, assuming it’s on five hours a day and that you’re paying the PUD’s retail rate of just over eight cents a kWh.

Again you say, “Meh.” I still don’t want to pay $50 for a bulb, even if it lasts longer and saves me money.

Now let’s expand our focus again. We’ll look at all 300,000 PUD customers to see the cumulative effect of everyone using the new Phillips bulb instead of incandescents. (By the way, the Phillips bulb is superior to existing LED and CFL bulbs in color rendition and sustained output over time, factors relevant to receiving the award.)

If the PUD bought 3 million Phillips bulbs it would cost $150 million. Buying five times as many incandescent bulbs would run $7,950,000 (assuming 53¢ ea.). But the savings lie in electricity consumption.

This may seem odd, but if the PUD (or its customers) invested $142 million in Phillips bulbs (over and above what the incandescent bulbs would cost) it would save a net of $173 million. That’s because customers would pay only $63 million for electricity (over the life of the Phillips bulb) whereas they would pay $378 million if they all continued to use 60-watt incandescent bulbs.

How difficult it would be for the utility to convince all 300,000 customers to buy the Phillips bulbs. Not going to happen. But why couldn’t the PUD buy the bulbs on behalf of its customers and figure out a way to ensure that they’re all installed?

After all, one way of thinking about a public utility is that it is the customers. Everything it does, from buying wholesale electricity to maintaining its fleet, is for customers. In exchange, customers pay a rate for each unit of electricity they consume. So, if the PUD bought all those bulbs the costs would be paid by the customers, just as they pay for the utility’s trucks and PUD employees’ desks.

In the long run all this makes sense. The customers—as a whole—invest the money and get a guaranteed return, expressed in lower electric bills. The PUD could borrow the money, and since interest rates are so low, the total debt service wouldn’t be much more than the initial capital.

The ThinkProgress article linked above mentioned that the Phillips bulb will last 30 years, which means that whoever ran the calculations assumed about two hours a day for the bulb, whereas I’ve used five hours to get a bulb-life of about 14 years. So, every 15 years, or 30, the PUD would sell low-interest bonds, buy millions of bulbs, and save the community millions of dollars a year.

Customers may think short term. The PUD doesn’t have to, and really shouldn’t. Indeed, it can afford to take the long view, which benefits its customers.

Still a long way to go before realizing The Goal. But I’ve described one place to start.

The strange Everett school board

Lots of ink has been spilled over the fighting (literal) amongst school board members, with member Jessica Olson playing the instigator—or so it seems. Last night the board took a bizarre step toward further inanity: limiting the review of the district’s superintendent to a committee of three; the board has five members. We may presume that Ms. Olson will not be invited to participate. She said that the 3-1 vote cemented the board’s “laughingstock” reputation, according to the Everett Herald.

For whatever reasons, she has given black marks to Superintendent Gary Cohn. But her review has been in the form of a minority report, which evidently didn’t become part of the record.

Now I don’t know any of the school board members. Nor am I in a position to finger one or more as the cause of the bitter animus between them, although readers of the Herald would likely infer that Olson is chiefly responsible.

But she has a point.

Olson contended the parameters for Cohn’s review should take place in a public meeting.

Having been through such processes as a member of the Snohomish County PUD board, I am quite familiar with the state’s Open Public Meetings Act, which gives deference to the people’s right to know what their public bodies are up to. The act allows for executive-session exemptions (pdf)—evaluating the performance of a public employee being one.

However, and this is where I think Olson has it right, the evaluation criteria are a matter of public record. Only the discussion of whether or not the employee satisfied the criteria belongs in executive session.

I’ve been on contentious boards, for which I was largely responsible. It’s not much fun, certainly. I’m relieved that we PUD commissioners managed to work through our problems—with the help of a professional facilitator. We’re now actually governing. Fancy that.