Passing the stress test for energy efficiency

Promoters of the Energy Star efficiency standard got egg on their faces last year when the Government Accountability Office submitted fake products and companies for the Energy Star label. The EPA approved 15 out of 20 bogus applications. The problem was that the Energy Star rules at the time let manufacturers certify product performance themselves. There was no uninterested party verifying efficiency claims.

That all changed Jan. 1 with the initiation of Energy Star's third-party certification program. Now accredited certification bodies (CBs) verify claims — they've already approved 10,000 new products. About 300 labs have been authorized to test products for certification, up from the 100 or so labs greeting the new initiative in January.

“The first four months was a rapid learning curve, but this second phase is a little easier,” summarizes Carl Bloomfield, global business line director for energy efficiency and lighting at Intertek, a CB that also has its own approved testing facilities. “Is it perfect? No, but we're at about 80 to 90%. Questions remain about how to test uniformly, some specs still are being written, but from our point of view we're almost there.”

The new setup and accounting method goes far beyond the old EPA method under which the agency simply collected basic efficiency data from product manufacturers. Now product makers appear generally satisfied with the scope and intent of the program. But they're growing impatient waiting for the infrastructure that makes the certification program more familiar, comfortable, and flexible.

The EPA, for instance, updates its database of new products bearing the Energy Star label about twice a month; manufacturers would like to see the database updated daily. (EPA indicates this should happen in Q2/Q3.) But the big issue is the cost of testing, plus the fact that many earlier products will not be grandfathered.

Participants aren't too concerned about meeting universally consistent testing procedures. Doing so economically is another issue. Product makers understand the rationale for standardization and third-party testing by independent, supervised (STLs), or witnessed testing laboratories (WTLs). But certification costs will likely ratchet up for product development, testing, and verification.

A case in point are the lighting, and solid-state lighting (SSL) devices which are part of the recent, complex Energy Star Luminaires v1.0 spec. Lighting industry insiders say there has been some informal grumbling about testing costs associated with residential light fixtures and “luminaires” (LEDs and their fixtures) which largely capture the Energy Star spotlight these days. Many lighting projects are funded by city- and county-based efforts at illuminating public roads, stadiums, and parking garages. Taxpayers will ultimately absorb these costs.

Meanwhile, the high cost of LEDs equivalent to incandescent and fluorescent bulbs for residential use makes their application prohibitive. Adding the cost of certification testing to these already expensive lighting options won't help their economics, critics contend.

About the costs

The costs to certify a product today are not insignificant. And things could get even more expensive with yearly verification of Energy Star requirements, a measure also included in the new certification regime. So economics will probably push lighting manufacturers to submit only their most promising or successful offerings for Energy Star status. Expect to see the same tradeoffs in other categories (white goods, consumer electronics, commercial food service equipment) as EPA fashions tougher specifications.

Still, EPA says it planned Energy Star certification procedures keeping costs in mind. “There were three key things we did to limit the manufacturer's burden, says Kathleen Vokes, Lead for Program Integrity at EPA. “First, we leveraged existing programs to test for safety and sanitation. So people like Underwriters Laboratories (UL) now also offer energy efficiency testing and certification services for Energy Star.

Page 2 of 3

“Second, we allowed in-house labs to qualify products. Manufacturers can use their own testing labs if they are supervised and witnessed by CBs. In many cases, these labs are separate from the manufacturing facilities. CBs can oversee these labs to make sure they are competent, pretty much the same way an accreditation body would monitor an independent laboratory.”

Perhaps the most contentious issue is that of verification testing over the life of the product. “Yes, we do have a requirement for verification testing to make sure manufacturers continue to meet Energy Star requirements,” Vokes confirms. “It takes the form of off-the-shelf testing to make sure what was certified initially is what consumers are buying. For that, we limited the number of products subject to verification testing to 10% of products tested each year.”

Manufacturers gripe about these audits, despite EPA's best efforts to accommodate them. In addition to the initial expense, there's a time and logistics burden dealing with any unexpected results or issues outside the manufacturer's control. And of course, there's a lot of paperwork required.

The annual verification process begins with a CB randomly selecting and testing a given number of products from 10% of a company's certified product models. Thus if a CB has previously certified 1,000 products, it will test about 100 each year. The CB gets test units by buying them in the open market. It generally buys one test sample for a given product model, or up to four for some products that are qualified based on multiple units, wherein statistical methods are used to determine compliance. Strict rules of procurement apply to ensure random sampling.

Once tests are complete, CBs inspect the resulting reports from the labs that provide the test data to ensure the product is still being manufactured in a manner consistent with initial testing and qualification. If the product passes, the CB will notify the manufacturer within days, as well as the EPA, which then places the product on its approved list (database).

The manufacturer gets a bill for testing and procurement, independent of the test results. Manufacturers will be given a short time to respond if a product sample fails to comply. But the penalties for failure are harsh: The product is “de-listed” and essentially deleted, in which case the manufacturer may need to start its Energy Star qualification from scratch.

EPA and manufacturers are working closely to avoid this scenario. One positive result of this close collaboration is that manufacturer (first-party) labs are likely to become more calibrated over time according to the steps/standards EPA wants to see in test procedures. That could potentially ensure that test results coming from first-party labs are virtually equivalent to those from the independent lab that passes final judgment.

Making the grade

But before manufacturers can worry about annual verifications, they first must get their initial certifications. CBs say some lighting manufacturers are chagrined to find out how long it takes to do so. “When it comes to lighting, the biggest issue is life-testing,” says Jason Chesley, Strategic Account Manager at CSA International's Alpharetta, Ga. operation. CSA is one of the country's major CBs and tested the entries in the Dept. of Energy's Bright Tomorrow Lighting Prize (L-Prize) competition.

Meeting in-situ temperature requirements and case temperature specs on the LED driver are related, often critical issues. SSL qualification is a long process done in stages, reportedly taking about two years to qualify a lamp at 25,000 hours (the minimum requirement for SSLs in residential applications). In comparison, it might take only five to ten business days to qualify computers and PCs, assuming they've been previously certified for safety. Commercial food service equipment such as grills can be certified in two to four weeks.

EPA allows interim qualification to 3,000 hours on the test rack. But lamps often start shifting color or dying early in that cycle. “If the manufacturer has, say, LM-80 (complete luminare test spec) data on the LEDs, we (the CB) can submit for early approval for Energy Star so they can do their initial marketing and sell their products,” says Chesley. “But it's a pending (early initial) qualification. The manufacturer must put the product back on the life rack and go from 3,000 to 6,000 hours to complete the initial life cycle. That's about a 10-month process. So it requires 6,000 hours (of LM-80 data) for a 25,000-hour lifetime claim.”

Then the product must burn in for 7,500 hours (for a 30,000-hour lifetime claim) and for 8,700 hours to make a (certified) 35,000-hour lifetime claim. The latter is the minimum requirement for commercial SSL devices. The bulb would have to go for 10,000 hours for a 40,000-hour device, and 12,500 hours for a 50,000-hour device.

The percentage of lamps that fail the test is close to 20% by some reports; that percentage could well be worse for some models that were on the shelves before the recent Energy Star initiative. “We've seen two major causes of the problem,” says Chesley, “poor construction (cold solder joints), and heat. If you don't dissipate the heat properly, the LED or driver will burn out.” He also sites rapid-cycle (on-off) testing as another pesky problem.

Waste not

The amount of testing that Energy Star requires is so significant that it engenders environmental issues. A modern lamp weighs perhaps two pounds and comprises of aluminum, packaging, and plastics. Lamps tested by the thousands and perhaps tens of thousands annually in the name of Energy Star eventually end up “in the trash.” Will the new requirements for testing and verification generate a significant disposal problem?

Page 3 of 3

Both users and observers think this might indeed be the case. The irony of consuming significant amounts of resources in the effort to save the environment is not lost on regulators. But presently there's no real alternative to life testing. Prevailing wisdom favors taking the long view towards energy conservation.

Concerns about end-of-life disposition aren't just a moral issue. EPA's most recent Luminaires v1.0 spec also directs that lighting products comply with the requirements of the ROHS directive for hazardous substances.

However, it looks as though individual states may have more to say about product make-up than Energy Star mandates. California first put legal requirements into place in 2007 through its Dept. of Toxic Substances Control for electronic devices such as computers, laptops and TVs with CRTs or LCDs. The Golden State now restricts the sale of general purpose lighting products containing such elements as mercury, lead, cadmium, and chromium with a law that became effective in January 2010. Other states are enacting similar measures. Ultimately, lighting manufacturers will be expected to know the source of each material in their products and will have to tell users how to dispose of them.

Resources

CSA International, http://www.csa-international.org

Dept. of Energy, http://www.energy.gov/energyefficiency/energystar.htm

Energy Star, http://www.energystar.gov/

Intertek, Arlington Heights, Ill., http://www.intertek.com

L-Prize, www.lightingprize.org

Lighting Facts, www.lightingfacts.com

Lighting Science Group Corp., Satellite Beach, Fla., http://www.lsgc.com/

Next Generation Luminaires, http://www.ngldc.org

EPA vs. the DoE

Energy Star isn't the only efficiency standard that manufacturers need to worry about. The Dept. of Energy (DoE) has efficiency standards in place for product areas that include home appliances, building materials, transportation, and others. Unfortunately, the efficiency picture can seem muddy to the uninitiated in that Energy Star categories overlap some of those covered by DoE standards. But products that are called on to meet a DoE qualifying standard by definition don't need to have a preexisting Energy Star label.

That's because the EPA-administered Energy Star program is voluntary and essentially sets a top-tier goal for energy efficiency standards. On the other hand, the corresponding DoE standards, usually mandatory, set the basic benchmark requirements.

In practice, Energy Star products often exceed DoE requirements by about 20%. Occasionally, technologies evolve to where the top-tier performance that Energy Star represents is pretty close to the minimum performance state-of-the-art technology can deliver. At this point no manufacturer gains by falling back to previous de-facto efficiency standards. Indeed, when this happens, the Energy Star standard can be “sunset,” as most recently happened last year with the EPA's retiring of external power supply (EPS) specifications.

Manufacturers covered by DoE efficiency standards should take heed: The agency is reportedly also looking at a third-party approach resembling that for Energy Star to handle its own market surveillance. One of the major reasons is that DoE and EPA standards don't completely overlap in all product categories. One example: motors.

Lighting directives and incentives

The DoE's voluntary new Lighting Facts program will showcase LED products from manufacturers that commit to testing procedures and reporting results according to strict industry standards (in this case, one called LM-79). The program is designed to ensure the LED products meet customer expectations for performance with respect to photometrics and color qualities (but not lamp lifetime).

In a sense, Lighting Facts takes over from DoE's original (and still existing, but modified) CALiPER program, which at that time served to qualify independent labs working on LED products. Lighting Facts products include LEDs for general lighting such as table lamps and outdoor light fixtures (but not flashlights or nightlights). Lighting Facts also plans to subject these products to verification testing once the program gets up to speed.

The DoE does not endorse products with a Lighting Facts label. However, the Energy Star program does require a Lighting Facts label be carried on certain items such as integral LED lamps for products not affected by the FTC labeling requirements taking effect Jan. 1, 2012.

DoE also sponsors the L Prize (i.e., Bright Tomorrow Lighting Prize), whose roots go back to the Energy Independence and Security Act (2007). The L Prize competition aims to find the best SSL products able to replace traditional 60-W incandescent lamps and PAR 38 halogens. The competition for the incandescent category calls for a lamp that consumes 10 W or less yet can deliver at least 900 lumens (i.e, 90 lumens/W) at a color temperature of 2700-3000 K, color rendering index of 90, and a lifetime of at least 25,000 hours.

The halogen category is presently on hold as the DoE modifies the requirements. In addition, the full requirements for a third category — the 21st Century Lamp, which can deliver at least 150 lumens/W — have yet to be developed. The DoE initially announced that up to three winners could be named in each category, with the first to meet all requirements for the 60-W category receiving a cash prize.

Philips Electronics was the first to enter a candidate in the 60-W category back in 2009, initially evaluated under the LM-79-08 test specifications and now in extended testing at Pacific Northwest National Laboratory (PNNL). Most recently, Lighting Science Group Corp. said it will submit its 60-W candidate (which it developed with Light Prescriptions Innovators LLC) to the DoE for initial testing.

No official end date to this competition has been verified, but there is still a lot of room for improving these devices. Indeed, DoE, which partnered with the Illuminating Engineering Society of America and the International Association of Lighting Designers to create the Next Generation Luminaires (NGL) competition, just announced more than 40 best-in-class winners at its third annual (2010) gathering.