Turning Off T12 Lighting… For Good!mn.gov/commerce-stat/pdfs/card-t12-lighting.pdf · vapor in a...

COMM-20150504-104138| March 31, 2017

Turning Off T12 Lighting… For Good!

A Market Characterization and Conservation

Potential Study

Conservation Applied Research & Development (CARD)

FINAL REPORT

Prepared for: Minnesota Department of Commerce

Division of Energy Resources

Prepared by: Franklin Energy Services, LLC

Prepared by:

Joe Plummer Dean Laube

Franklin Energy 2303 Wycliff St., Suite 2E St. Paul, MN 55114 Phone: (612)284-3663 website: Franklin Energy website (www.franklinenergy.com)

© 2017 Franklin Energy Services, LLC. All rights reserved.

Contract Number: 104138

Prepared for Minnesota Department of Commerce, Division of Energy Resources

Mike Rothman, Commissioner, Department of Commerce

Bill Grant, Deputy Commissioner, Department of Commerce, Division of Energy Resources

Mark Garofano, Project Manager (651)539-1864 [email protected]

ACKNOWLEDGEMENTS

This project was supported in part by a grant from the Minnesota Department of Commerce, Division of Energy Resources, through the Conservation Applied Research and Development (CARD) program, which is funded by Minnesota ratepayers.

The authors would also like to thank the following individuals for their contributions to this project: Eileen Hannigan of ILLUME Advising, LLC for her assistance with the experimental design and data analysis; Allen Anderson of Franklin Energy Services for sharing his lighting expertise and collecting field data; and Paul Bertucci, Mark Francis, Brandon Johnson, Daniel Martin, and Jacob Semann of Franklin Energy Services for collecting data from over 200 business locations across the state. Last but not least, we are grateful to the business owners and employees that graciously agreed to participate in this study.

DISCLAIMER

This report does not necessarily represent the view(s), opinion(s), or position(s) of the Minnesota Department of Commerce (Commerce), its employees or the State of Minnesota (State). When applicable, the State will evaluate the results of this research for inclusion in Conservation Improvement Program (CIP) portfolios and communicate its recommendations in separate document(s).

Commerce, the State, its employees, contractors, subcontractors, project participants, the organizations listed herein, or any person on behalf of any of the organizations mentioned herein make no warranty, express or implied, with respect to the use of any information, apparatus, method, or process disclosed in this document. Furthermore, the aforementioned parties assume no liability for the information in this report with respect to the use of, or damages resulting from the use of, any information, apparatus, method, or process disclosed in this document; nor does any party represent that the use of this information will not infringe upon privately owned rights.

http://www.franklinenergy.com/

i

Contents

Executive Summary ................................................................................................................................... 1

Introduction ................................................................................................................................................ 3

Background and Objectives .................................................................................................................. 3

A Brief History of Commercial Lighting Standards.......................................................................... 4

Methodology ............................................................................................................................................... 5

Sample Design ........................................................................................................................................ 5

Site Assessment Process ........................................................................................................................ 8

Data Validation ....................................................................................................................................... 9

Assumptions ........................................................................................................................................... 9

Findings ..................................................................................................................................................... 11

T12 Prevalence ...................................................................................................................................... 11

T12 Lighting Load ................................................................................................................................ 12

T12 Energy Consumption ................................................................................................................... 13

T12 Lamp and Fixture Characteristics .............................................................................................. 14

T12 Average Power Density ............................................................................................................... 15

Energy Savings Potential .................................................................................................................... 16

Discussion of Results ............................................................................................................................... 24

Key Conclusions ................................................................................................................................... 24

Programmatic Approaches ................................................................................................................. 25

Replacement Product Selection ...................................................................................................... 25

Targeted Outreach ........................................................................................................................... 28

Incentives, Buydowns, and Copays .............................................................................................. 29

Messaging ......................................................................................................................................... 30

TRM Recommendations ...................................................................................................................... 30

Conclusion ................................................................................................................................................ 33

References ................................................................................................................................................. 34

Appendix A. Data Collection Template ................................................................................................ 36

Appendix B. Derivation of Statewide Lighting Load ......................................................................... 37

Appendix C. Fixture Lookup Table ....................................................................................................... 38

Appendix D. Supplemental T12 Data ................................................................................................... 39

ii

List of Figures

Figure 1. Map of Sample Regions. ........................................................................................................... 6

Figure 2. Statewide Energy Savings Potential at Varying Incentive Levels. ................................... 20

Figure 3. Lifetime Savings Example for LED Kit. ................................................................................ 31

List of Tables

Table 1. T12 Prevalence by Business Size. .............................................................................................. 1

Table 2. Results of Power Analysis .......................................................................................................... 5

Table 3. Area, Population, and Business Count by Region. ................................................................. 5

Table 4. Business Size Classifications. ..................................................................................................... 7

Table 5. Selected Communities in Each Region. .................................................................................... 7

Table 6. Assumed (Default) T12 Lamp Wattages. ............................................................................... 10

Table 7. Number of Sites with T12s by Region and Size Category.. ................................................. 11

Table 8. T12 Load by Region and Size................................................................................................... 12

Table 9. T12 Energy Consumption by Region and Size Category. ................................................... 13

Table 10. Average Lighting Hours in Study Compared to Deemed Hours in Minnesota TRM v2.0.. ........................................................................................................................................................... 14

Table 11. T12 Fixture Distribution by Lamp Length. .......................................................................... 14

Table 12. Distribution of T12 Lamp Quantity per Fixture. ................................................................. 15

Table 13. T12 Average Power Density by Region and Size Category. ............................................. 15

Table 14. Technical, Economical, and Achievable Energy Savings Potential by Region and Size Category. ................................................................................................................................................... 18

Table 15. Statewide Technical, Economical, and Achievable Savings Potential and Projected Costs. .......................................................................................................................................................... 18

Table 16. Statewide Energy Savings Potential at Varying Incentive Levels. ................................... 19

Table 17. Equivalent Dollar Amounts for Modeled Incentive Levels. ............................................. 21

Table 18. Annual Operating Hours Necessary for a Two-Year Payback or Less............................ 22

Table 19. Summary of T12 Replacement Options................................................................................ 27

Turning Off T12 Lighting COMM- 104138 | March 2017 Franklin Energy 1 | P a g e

Executive Summary

Although US federal standards have effectively phased out the manufacture and import of fluorescent T12 lamps and ballasts, this outdated technology lives on. Businesses that continue to operate T12 lighting systems are burdened with an inefficient lighting source with inferior lighting quality and performance.

Utilities throughout Minnesota and North America have worked for years to help customers upgrade their T12s, which historically meant converting to more efficient T8 or T5 systems; however, not all T12s have been converted. Having a reliable estimate of remaining T12 stock would be useful for Conservation Improvement Program (CIP) planning.

This study set out to quantify remaining T12 stock in Minnesota through field assessments of more than 200 randomly-selected businesses spread across a random sample of 10 communities. Analysis of this data has confirmed what many observers suspected: T12 lighting is far from gone. We estimate that one-in-four non-residential buildings, statewide, have at least one T12 fixture, accounting for 242 MW in total, approximately 10 percent of the total commercial/industrial lighting load. On an energy basis, these fixtures consume 881 GWh per year, equivalent to the annual consumption of approximately 77,000 US homes.1

Past efforts by utilities to incentivize T12 replacements have generally relied heavily on bonus offers. Future efforts will likely require targeted outreach to influence remaining customers with T12s to convert their working fixtures. Our results show that business locations of 5,000 square feet or less, the most numerous size category in the state, have the highest average power density of T12 fixtures, as shown in Table 1.

Table 1. T12 Prevalence by Business Size. Research indicates that small businesses under 5,000

square feet present the highest average power density of T12 fixtures.

Size Range Business Count Average T12 Watts/ft2

< 5,000 ft2 104,639 0.284

5,000 – 9,999 ft2 41,905 0.116

≥ 10,000 ft2 91,875 0.043

The high prevalence of T12s in small businesses presents multiple challenges to the typical CIP program; however, one solution that has been deployed successfully in Minnesota is community-based small business energy “blitzes,” in which teams of utility representatives canvass a community, performing quick commercial energy audits focusing on the top three to four efficiency opportunities in each building. Because small businesses tend to be clustered in central business districts, this approach allows representatives to reach many locations in a relatively short time span. Focusing on the top three to four opportunities with the best paybacks, which usually includes lighting, reduces the cost and time of each audit, while not overwhelming the business owner with a lengthy list of recommendations. These campaigns

1 Based on an average electricity consumption of 11,300 kWh per year, derived from the US Energy Information Administration’s 2009 Residential Energy Consumption Survey.


not only create awareness of energy efficiency opportunities and matching incentives, but generate goodwill for the utility. Timing the campaign to coincide with a special bonus offer or “bounty” for T12 replacements can be an especially effective strategy.

Another approach that has proven effective is to leverage local trade allies via a midstream incentive program, in which incentives are offered to lighting contractors for promoting T12 upgrades. Small businesses often have long-standing, trusted relationships with local contractors and rely on their expertise when making purchasing decisions. Through focused outreach, education, and trade incentives, trade allies can be motivated to identify and promote T12 conversions.

The report concludes with a discussion of considerations for treatment of T12s in future versions of the Minnesota Technical Reference Manual (TRM). Based on our findings, we recommend that Commerce consider modifying the existing T12 retrofit measures to allow utilities to claim full savings for direct conversion to Light-Emitting Diode (LED) technologies. Furthermore, given the uncertainty over the remaining useful life of current T12 fixtures, coupled with current fluorescent lighting standards, we recommend that Commerce consider a one-year deemed measure lifetime for T12 to LED conversions.


Introduction

Background and Objectives

Not long ago, T12 fluorescent lighting systems were the standard lighting technology in commercial buildings. This base has gradually eroded over time as customers, aided by utility demand-side management (DSM) programs, upgraded to more efficient T8 and T5 systems. Today, the commercial lighting market is in the midst of a remarkable transformation. LED products have rapidly gained market share, offering better performance and longevity than fluorescent technologies. However, in our experience, it is still common to find T12 fixtures installed throughout Minnesota, especially in small businesses.

Businesses that continue to operate T12 lighting systems are encumbered with an inefficient lighting source providing inferior lighting quality and performance. Although these businesses could realize significant savings through upgrading to modern fluorescent or LED systems, they are often budget-constrained and/or lack awareness of cost-effective replacement solutions.

Utilities throughout Minnesota and North America have worked for years to help customers upgrade their T12s, which historically meant converting to more efficient T8 or T5 systems; however, not all T12s have been converted. Having a reliable estimate of remaining T12 stock would be useful for CIP managers as they allocate program budgets toward achievement of energy savings goals.

With support from the Minnesota Department of Commerce, Division of Energy Resources (“Commerce”), this study set out to quantify the remaining T12 stock in Minnesota, a state with more than 30 years of DSM history. This was accomplished by gathering a statistically representative sample of fixture and building data from over 200 randomly selected businesses spread across a random sample of 10 Minnesota communities.

Drawing from this rich dataset, we developed estimates of both the total energy T12 fixtures consume statewide and the potential energy savings that could be realized by converting these fixtures to LED systems. We also offer insight as to how CIP programs can capture this energy savings potential through targeted outreach and incentives. Finally, we provide recommendations regarding the continued treatment of T12 measures in the Minnesota TRM.


A Brief History of Commercial Lighting Standards

Fluorescent lighting is a form of gas discharge lighting. All gas discharge lighting operates on the same basic principle in which an electric current is passed through a low-pressure mercury vapor in a glass tube. The excited mercury atoms emit ultraviolet light, which strikes a phosphor coating on the inside of the glass tube, causing it to fluoresce, or emit visible light.

The first federal standards for general service fluorescent lamps were enacted by Congress in the Energy Policy Act of 1992 (EPACT). These standards were updated by the US Department of Energy (DOE) in 2009 and took effect on July 14, 2012. The amended standards set minimum efficacy (lumens per watt or “LPW”) requirements for general service fluorescent lamps equivalent to Series 800 T8s, effectively phasing out the manufacture or import of general service T12 lamps. At that time, however, due to a shortage of certain rare earth oxides used to produce phosphor coatings, some manufacturers were granted exception relief through July 2014 (ASAP, 2017).

Federal standards have also been enacted for fluorescent ballasts. Every fluorescent light fixture requires a ballast, which serves to initiate and regulate the electric current that flows through the lamps. The first standards for fluorescent ballasts were established as part of the National Appliance Energy Conservation Act (NAECA) of 1990, which effectively outlawed the manufacture or import of what are referred to as “standard” magnetic ballasts. All magnetic ballasts produced after this date are termed “energy saving” or “energy efficient” magnetic ballasts (NLPIP, 1993).

The ballast standards were amended in 2000 by the DOE. The 2000 Rule effectively phased out the inclusion of magnetic T12 ballasts in new fixtures manufactured or imported after July 1, 2005, and, by July 1, 2010, phased out the manufacture or import of all T12 magnetic ballasts. There are T12 electronic ballasts that meet the efficiency requirements of the 2000 Rule; however, it is rare to find these in the field. New standards were adopted by DOE in 2011 that expanded the scope of coverage for ballasts to include T8 and T5 systems, among other categories, beginning in 2014 (ASAP, 2017).

There are no mandatory standards for LED lighting systems. However, there are voluntary standards for LED screw-in bulbs through ENERGY STAR and for commercial LED products through Design Lights Consortium (DLC). Many utilities require, or at a minimum recommend, the use of ENERGY STAR and DLC standards for rebate eligibility.


Methodology

Sample Design

To achieve a high degree of accuracy while minimizing cost, this study uses a cluster sample design. Under this framework, selected businesses were clustered in several communities spread across the state to minimize travel time.

With the goal of quantifying T12s with a margin of error of less than 10 percent at a confidence level of 90 percent, we used a power analysis to evaluate different clustering options. The power analysis assumed some correlation between T12s and the communities in which they are located. This reduces the equivalent sample size and increases the margin of error. It is reasonable to expect some correlation because of underlying factors such as the overall economic health of the community or the effectiveness of the incumbent utility’s CIP activities. Error! Reference source not found.Table 2 provides the results of the power analysis. Test Case 4 was chosen as the best balance of accuracy with cost.

Table 2. Results of Power Analysis. Including 10 communities and 20 businesses per community,

with an 8% margin of error at 90% confidence level, Test Case 4 provided the best balance of

accuracy and cost.

Test Case

Number of Communities

Visited

Number of Businesses per

Community Visited

Total Sample

Size Design Effect

Equivalent Sample

Size

Margin of Error (%) at

90% Confidence

Level

1 40 5 200 1.2 167 6

2 30 7 210 1.3 162 6

3 20 10 200 1.45 138 7

4 10 20 200 1.95 103 8

5 5 40 200 2.95 68 10

Table 3. Area, Population, and Business Count by Region. Excluding the Metro region, the study

sized regions to include similar populations and business counts.

Region Square Miles Population

Total Business

Count Selected

Communities Selected

Businesses

Metro 2,811 2,985,405 124,015 2 40

Northeast 24,603 581,820 27,081 2 40

Northwest 24,990 576,697 25,157 2 40

Southeast 9,106 627,295 29,457 2 40

Southwest 18,100 685,956 32,709 2 40

Total 79,610 5,457,173 238,419 10 200

To provide localized results to inform CIP programming, the sample was stratified across five geographic regions: the Northwest (NW), Northeast (NE), Southeast (SE), Southwest (SW), and


the seven-county metropolitan area2 encompassing Minneapolis, St. Paul, and the surrounding suburbs. We selected two communities within each region, with a sample size of 20 businesses per community. With the exception of the metro region, we sized each region to contain roughly equivalent populations and business counts, as indicated in Table 3.

Figure 1Error! Not a valid bookmark self-reference. provides a map of the sample regions. Note that the region boundaries followed county lines—this approach enabled the use of county-level statistics provided by the US Census Bureau and the Salesgenie® commercial marketing database, described below. The selected communities in each region are indicated on the map and are listed in Table 5. The sample communities provide a wide range of sizes from which to capture data.

Figure 1. Map of Sample Regions. Using county lines enabled the use of valuable US Census

Bureau and Salesgenie data

To ensure that a range of business sizes was captured in the sample, we further stratified the sample by business size, as measured by floor area. Three size classifications were defined, as

2 Includes Anoka, Carver, Dakota, Hennepin, Ramsey, Scott, and Washington counties.


shown in Table 4. A minimum quota of six sites per size range, and 20 sites per community was required.

Table 4. Business Size Classifications. To ensure data capture across a range of business sizes, the

study included a minimum quota of six sites per size range.

Size Square Footage

Sample Count per

Community

Small < 5,000 6-7

Medium 5,000 - 9,999 6-7

Large ≥ 10,000 6-7

Total 20

Table 5. Selected Communities in Each Region. The sample communities provided a wide range of

sizes from which to capture data.

Region Community Business

Count

Metro3

Minneapolis 57,897

St. Paul 33,742

Northeast

Duluth 7,204

Nisswa 311

Northwest

Alexandria 1,773

Foley 243

Southeast

Albert Lea 1,135

Red Wing 1,082

Southwest

Morris 501

Worthington 860

We generated the sample set in two stages. In the first stage, two communities per region were randomly selected after applying a probability weighting method (WHO, 2016), in which weights were assigned to each community in a region based on business count. This method increased the probability of selecting larger communities (as measured by business count). We did this to compensate for the fact that small cities are much more numerous than large cities. If purely random sampling had been used, the sample would have been heavily skewed toward small cities, altering the viability of the resultant data.

Because many small towns in Minnesota do not contain even 20 businesses, cities with fewer than 40 businesses were first removed from the dataset. (Forty was chosen to provide a safety margin to ensure that 20 complete site samples could be collected in each city.) This removal has no practical impact on the study findings.

We implemented the selection process in Microsoft® Excel® using random number generation functions. Salesgenie was queried for the business count in each community, which was needed

3 The Minneapolis and St. Paul business counts include businesses that are actually located in a suburb, but whose location address was specified as Minneapolis or St. Paul in Salesgenie.


to compute the probability weighting factors. Because this study is focused on commercial buildings, the queries excluded home-based businesses.

The final communities selected in each region and corresponding business counts are summarized in Table 5. Note that even though the selection process was weighted toward larger communities, the overall sample contains a wide range of community sizes.

The second stage of the sample generation process was to randomly select businesses for site assessments. For this step, a total of 60 businesses, including 20 businesses in each size range (Small, Medium, and Large), was randomly selected from Salesgenie queries based on square footage (excluding home-based businesses). A list of 60 businesses was generated per city to provide a safety margin to ensure that field staff would be able to complete at least 20 site assessments.4

This process resulted in a list of business names and addresses partitioned by size range in each of the 10 sample communities. Field staff referenced these lists for site selection in each community. Each staff person was given a quota of 6 to 7 completions in each size range, resulting in at least 20 completions per community.

Site Assessment Process

Site assessments were completed throughout 2016. These assessments were largely a “cold-calling” activity. For some communities, we mailed postcards to the selected businesses ahead of time to notify them of the upcoming visit and give an overview of the study; however, because of the dynamic nature of our daily business activities, there was not always sufficient lead time to send postcards. Fortunately, not receiving a postcard had no perceivable impact on whether a business agreed to participate in the study. Even when a postcard was sent, the actual employee(s) present when field staff arrived often had not seen the postcard.

To encourage participation, the study initially offered two ENERGY STAR-rated LED bulbs to each business in exchange for its participation. However, as the study progressed, we found that most businesses were indifferent to the LED offer. The vast majority of sites agreed to participate. Those that did not agree to participate generally voiced privacy concerns. In other cases, an employee wanted permission from the owner, who was not present, prior to participating. Field staff carried an information sheet that provided an overview of the study and contact information for Commerce and the primary investigator in case of questions. We did find that the information sheet was helpful to bolster the study’s perceived credibility.

The site assessment protocol was designed to allow the assessment to be completed quickly and with minimal disruption to the business. After entering the premises and receiving permission to perform the assessment, staff walked through the building and collected the following information on a standardized form:

T12 fixture counts and lamp wattages

4 Ideally, the first stage of sampling, city selection, would have also used 60 businesses as the minimum community size.


T12 ballast types (magnetic or electronic)

Spare T12 lamp counts and wattages

Business hours

Fixture operating hours (provided by business)

A/C presence

We captured additional information to facilitate further analysis outside of this report. Refer to the form template in Appendix A for the information types collected.

Ballast discriminators were used to detect the type of each T12 ballast, magnetic or electronic, without having to disassemble the fixture. Information on operating hours was provided by employees or based on posted information; no data logging was done. If the employee knew the lamp wattages with certainty, staff recorded this information; otherwise it was left blank. Field staff were not required to inspect lamps to capture wattages as this would have been time-consuming, disruptive, and could create a liability for any accidental damage incurred.

For some sites, staff discovered that the square footage range from Salesgenie appeared to be incorrect. This occasionally happened for businesses located in multi-tenant buildings such as shopping malls; in these cases, the location size appeared to correspond to the entire building rather than the individual site. To mitigate this issue, field staff estimated and recorded the correct size classification (Small or Medium) when on-site. Because actual square footage was not needed, a high degree of accuracy was not necessary.

Once staff completed forms for each site, we developed a Visual Basic program to compile all the data into a single Excel sheet for analysis.

Data Validation

Once compiled, the data was carefully validated for accuracy and completeness. Our validation included removing typographical errors, checking for outliers, filling in missing information where possible, and removing invalid characters. We documented all data changes, and the original source files were left unaltered.

Assumptions

Because not all data points needed for the analysis could be observed, some values had to be assumed. With regard to lamp wattage, the following method was used when the actual wattage was unknown:

1. If spare T12 lamps were found on-site, the spare lamp wattage was used for the installed lamp wattage. For example, if spare T12 4-foot 34W lamps were found in a storeroom, then all installed T12 4-foot lamps were assumed to be 34W.

2. If no spare T12 lamps were found on-site, then the average observed wattage for each lamp type in the sample was used. For example, the average observed 4-foot T12 wattage from the sample was 42.4W. Because 40W is the closest wattage that is commercially available, 40W was assumed for 4-foot T12s of unknown wattage.


Similarly, for 8-foot T12s, a mix of 60W and 75W lamps was observed. The average observed wattage was determined to be 73.1. Therefore, for 8-foot T12s of unknown wattage, 75W was assumed.

Only 2 percent of the 1,650 total T12 lamps observed in the sample data were under 4 feet in length, and no actual wattages were observed. Therefore, for simplicity, T12s under 4 feet were assumed to be 20W, half the assumed wattage for 4-foot lamps.

Table 6 provides the relative proportion of each lamp length observed in the sample data and corresponding assumed wattage.

Table 6. Assumed (Default) T12 Lamp Wattages. Where actual wattage values were unavailable,

the study used the justified assumptions above.

Lamp Length

Proportion of Sample

Assumed Wattage

< 4-foot 2% 20 W

4-foot 82% 40 W

8-foot 16% 75 W

The total wattage of a fluorescent fixture is a function of the number of lamps, the lamp wattage, and the type of ballast. As confirmed by our field data, the vast majority of T12 fixtures use magnetic ballasts; however, the ballast discriminators were not able to distinguish between standard magnetic ballasts and energy saving magnetic ballasts. Therefore, to be conservative, all magnetic ballasts were assumed to be the energy saving type, which are approximately 8 percent more efficient than standard magnetic ballasts.5

For calculating average T12 power densities (Watts per square foot) by size category, it was necessary to assume an average square footage for each site because Salesgenie only specifies a square footage range. We used the following approach:

1. For Small and Medium buildings, we assumed average square footages of 2,500 and 7,500, respectively, equal to the midpoint of each size range (Small = 0 to 4,999 square feet, Medium = 5,000 to 9,999 square feet).

2. For Large buildings, which do not have a defined upper bound for square footage

(Large = 10,000 ft2 or greater), we derived an average square footage by first subtracting

the total square footage of Small and Medium buildings, derived from the average

square footages in Minnesota from Salesgenie, from described in Step 1 and the tally of

Small and Medium businesses in the total square footage of C/I buildings in Minnesota

as calculated in Appendix B. The resulting square footage for Large businesses was then

divided by the tally of Large businesses from Salesgenie to yield an average square

footage of 32,777.

5 Based on analysis of deemed fixture wattages in Appendix B of the Minnesota TRM.


Findings

T12 Prevalence

Of the 210 buildings randomly selected for site assessments, 56 buildings (approximately one in four) contained T12 fixtures. Applying statewide weighting factors, we estimate that 28 percent of non-residential buildings in Minnesota have at least one T12 fixture. Regional level results are presented in Table 7.

Table 7. Number of Sites with T12s by Region and Size Category.

Region

Sample Count

Sites with T12s,

Sample

% with T12s,

Sample Sub-set

Population6 Sub-set

Statewide Weighting

Factor

% with T12,

Statewide Size

Metro 43 13 30.2% 124,015 0.520 15.7%

Small 14 4 28.6% 46,576 0.195 5.6%

Medium 15 5 33.3% 23,105 0.097 3.2%

Large 14 4 28.6% 54,334 0.228 6.5%

Northeast 41 15 36.6% 27,081 0.114 4.2%

Small 18 9 50.0% 13,564 0.057 2.8%

Medium 14 3 21.4% 4,629 0.019 0.4%

Large 9 3 33.3% 8,888 0.037 1.2%

Northwest 45 6 13.3% 25,157 0.106 1.4%

Small 16 3 18.8% 13,797 0.058 1.1%

Medium 13 1 7.7% 4,276 0.018 0.1%

Large 16 2 12.5% 7,084 0.030 0.4%

Southeast 41 11 26.8% 29,457 0.124 3.3%

Small 13 5 38.5% 13,477 0.057 2.2%

Medium 13 4 30.8% 4,464 0.019 0.6%

Large 15 2 13.3% 11,516 0.048 0.6%

Southwest 40 11 27.5% 32,709 0.137 3.8%

Small 17 7 41.2% 17,225 0.072 3.0%

Medium 11 3 27.3% 5,431 0.023 0.6%

Large 12 1 8.3% 10,053 0.042 0.4%

Total (Avg.) 210 56 (26.7%) 238,419 1.000 28.4%

6 Throughout this section, “population” refers to total business count. The population figures were sourced from Salesgenie.


T12 Lighting Load

Regional and statewide T12 load was derived from the average T12 load (kW) per site for each region and size category. We then calculated the total T12 load for each subset by multiplying by the population size (number of businesses). As indicated in Table 8 we estimate the total T12 load statewide is 241,805 kW (242 MW).

Table 8. T12 Load by Region and Size. Results show a T12 statewide load of 242 MW.

Region

Sample Count

T12 Load (kW)

Avg. kW per Site

Population Sub-set

Total T12 Load (kW),

Population Sub-set Size

Metro 0.964 124,015 134,749

Small 14 7.6 0.540 46,576 25,131

Medium 15 9.4 0.625 23,105 14,441

Large 14 24.5 1.752 54,334 95,178

Northeast 1.044 27,081 34,362

Small 18 21.5 1.194 13,564 16,190

Medium 14 4.4 0.314 4,629 1,454

Large 9 16.9 1.881 8,888 16,718

Northwest 0.331 25,157 8,019

Small 16 3.5 0.219 13,797 3,017

Medium 13 0.4 0.032 4,276 135

Large 16 11.0 0.687 7,084 4,867

Southeast 1.023 29,457 32,070

Small 13 16.3 1.252 13,477 16,879

Medium 13 10.6 0.815 4,464 3,640

Large 15 15.0 1.003 11,516 11,551

Southwest 1.256 32,709 32,605

Small 17 13.0 0.765 17,22 5 13,169

Medium 11 34.2 3.107 5,431 16,876

Large 12 3.1 0.255 10,053 2,560

Total (Avg.) (0.911) 238,419 241,805

To verify the calculated T12 load of 242 MW, we compared the Salesgenie total business count of 238,419 to the total number of commercial and industrial electric utility customers reported by Minnesota utilities to the US Energy Information Administration for 2015: 298,867 (EIA, 2016). Because the EIA figure is about 25% higher than the Salesgenie total, it is very possible that the actual T12 statewide load is higher than our estimate of 242 MW.

To put the total estimated load of 242 MW in context, we estimate the total Commercial/Industrial (C/I) lighting load statewide is approximately 2,558 MW (refer to Appendix B). Therefore, T12 lighting accounts for nearly 10 percent of the statewide total C/I lighting load in Minnesota, a state with over 30 years of DSM history. The proportion of T12s in a state with minimal DSM activity may be significantly higher.


T12 Energy Consumption

Using a similar process, we calculated the energy consumed annually by T12 fixtures with the recorded hours of operation for each site. We included an HVAC factor in the calculation to account for the impact of lighting power on air-conditioning load, where present, to maintain consistency with the TRM methodology.7 The results of this process are summarized in Table 9 and indicate that the total energy consumed by T12 fixtures, statewide, on an annual basis is approximately 881 GWh at the meter. This is approximately 11 percent of the total lighting energy consumed in C/I buildings statewide (refer to Appendix B).

Table 9. T12 Energy Consumption by Region and Size Category. Results indicate a total T12

consumption of 881 GWh statewide.

Region Sample Count T12 kWh

Avg. kWh per Site

Population Sub-set Total T12 GWh Size

Metro 43 159,923 3,719 124,015 461.2

Small 14 21,336 1,524 46,576 71.0

Medium 15 71,796 4,786 23,105 110.6

Large 14 66,791 4,771 54,334 259.2

Northeast 41 137,536 3,355 27,081 90.8

Small 18 59,953 3,331 13,564 45.2

Medium 14 33,699 2,407 4,629 11.1

Large 9 43,884 4,876 8,888 43.3

Northwest 45 40,006 889 25,157 22.4

Small 16 11,863 741 13,797 10.2

Medium 13 467 36 4,276 0.2

Large 16 27,677 1,730 7,084 12.3

Southeast 41 131,719 3,213 29,457 94.6

Small 13 49,785 3,830 13,477 51.6

Medium 13 32,004 2,462 4,464 11.0

Large 15 49,930 3,329 11,516 38.3

Southwest 40 259,452 6,486 32,709 212.2

Small 17 34,852 2,050 17,225 35.3

Medium 11 206,081 18,735 5,431 101.7

Large 12 18,520 1,543 10,053 15.5

Total T12 Consumption (GWh) 881.2

By our estimates, this consumption level accounts for more than 600,000 metric tons of carbon dioxide emissions annually from the generation of electricity. This is equivalent to the annual emissions of over 130,000 cars (EPA, 2017).

On a statewide level, our data shows a weighted-average operating hours for T12 fixtures of 3,210 hours per year, or approximately 60 hours per week. The weighted-average operating hours for all fixtures equated to a similar figure, 3,226 hours per year. Considering the deemed

7 Refer to C/I Lighting End Use in the Minnesota TRM v2.0.


operating hours for C/I lighting in the MN TRM, this result was lower than expected (see Table 10).

Table 10. Average Lighting Hours in Study Compared to Deemed Hours in Minnesota TRM v2.0.

Study hours were lower than expected.

Building Type Hours

Elementary School 2,422

Hotel/Motel 3,044

Study Findings 3,226

College 3,540

Restaurant 3,673

Secondary School 4,311

Office 4,439

Other/Misc. 4,576

Retail 4,719

Warehouse 4,746

Exterior lighting 4,903

Health 5,095

Manufacturing 5,200

Grocery/Supermarket 5,802

Hospital 6,038

24-Hour Facility 8,766

Safety or Code Required 8,766

T12 Lamp and Fixture Characteristics

To inform ongoing program activities, we collected information during the study on the types of T12 fixtures most commonly found in Minnesota businesses. Table 11 shows the predicted distribution of lamp sizes after applying statewide weighting factors. As indicated, the majority (83 percent) of T12 fixtures have 4-foot lamps.

Table 11. T12 Fixture Distribution by Lamp Length.

Lamp type Proportion

4-foot 83%

8-foot 15%

Under 4-foot 2%

Table 12 shows the predicted distribution of T12 fixtures statewide by number of lamps per fixture. As indicated, the majority of 4-foot and 8-foot T12 fixtures have two or more lamps. Most T12 fixtures under 4 feet, which were found to be very rare (see Table 11), are expected to be one-lamp fixtures.


Table 12. Distribution of T12 Lamp Quantity per Fixture.

No. of Lamps

Lamp Type

4-foot 8-foot Under 4-

foot

1 8% 17% 79%

2 52% 83% 21%

3 7% 0% 0%

4 33% 0% 0%

T12 Average Power Density

A key consideration for energy efficiency program managers is likely to be where T12s are most likely to be found. We attempted to answer this question by calculating the average power density (Watts per square foot) of T12s in each location size category.

After calculating the T12 power density for each site as described in the Assumptions section under Methodology, we calculated average T12 power densities across each region and size category as shown in Table 13.

Table 13. T12 Average Power Density by Region and Size Category.

Region Population Sub-

set


Population Sub-set Avg. Sq Ft

Ave T12 Watts per

Sq Ft Size

Metro 124,015 134,749 16,697 0.065

Small 46,576 25,131 2,500 0.216

Medium 23,105 14,441 7,500 0.083

Large 54,334 95,178 32,777 0.053

Northeast 27,081 34,362 13,292 0.095

Small 13,564 16,190 2,500 0.477

Medium 4,629 1,454 7,500 0.042

Large 8,888 16,718 32,777 0.057

Northwest 25,157 8,019 11,876 0.027

Small 13,797 3,017 2,500 0.087

Medium 4,276 135 7,500 0.004

Large 7,084 4,867 32,777 0.021

Southeast 29,457 32,070 15,094 0.072

Small 13,477 16,879 2,500 0.501

Medium 4,464 3,640 7,500 0.109

Large 11,516 11,551 32,777 0.031

Southwest 32,709 32,605 12,636 0.079

Small 17,225 13,169 2,500 0.306

Medium 5,431 16,876 7,500 0.414


Region Population Sub-

set


Population Sub-set Avg. Sq Ft

Ave T12 Watts per

Sq Ft Size

Large 10,053 2,560 32,777 0.008

Statewide 238,419 241,805 15,046 0.067

Small 104,639 74,387 2,500 0.284

Medium 41,905 36,544 7,500 0.116

Large 91,875 130,874 32,777 0.043

The results of this process, indicate that on average, T12 are most prevalent in Small buildings in every region except the Southwest. T12s are also most prevalent on a statewide basis in Small buildings, followed in descending order by Medium and Large buildings. There is no apparent reason other than random chance in the specific cities and business locations that were sampled for why the Southwest region was an exception to this overall finding.

The general conclusion that Small business locations have the highest T12 power density on average is consistent with our anecdotal experience prior to this study. We speculate that economic factors – in our experience, small businesses are often cash-strapped, and cannot access volumetric pricing – and less CIP outreach historically (utility account managers typically have more direct interaction with larger businesses) are the underlying reasons for this finding.

To put the average power densities in Table 13 in context, the maximum lighting power density for an office building allowed under the current Minnesota energy code is 0.9 Watts per square foot (Minnesota Rules, 2015).

Energy Savings Potential

One of the objectives of this study was to estimate the technical, economical, and achievable energy savings potential for upgrading T12 lighting. The definition of these terms varies, but generally speaking, technical energy savings potential is understood to mean the maximum theoretical energy savings that could be realized based on best available technology, economic potential is the sum of cost-effective savings opportunities, and achievable energy savings potential is the savings that could be achieved under the most aggressive program implementation strategy.

For the purpose of this study, we defined each of these terms as follows:

Technical energy savings potential is the theoretical energy savings that could be achieved if every T12 luminaire in Minnesota was upgraded to the most efficient luminaire available, regardless of cost.

Economic energy savings potential is the savings that could be realized if rebates were set to 50 percent of incremental cost on average, and customers require a payback of two years or less. Although actual payback requirements vary by customer, two-year or less is a common range for many commercial businesses.


Achievable energy savings potential is defined as the total energy savings potential if rebates were set to 100 percent of incremental cost using average efficacies of DLC-qualified products.

The technical, economical, and achievable energy savings potentials from upgrading T12s were calculated on a site-by-site basis and then rolled up to the state level. Two broad categories of LED replacement options were included in the analysis: retrofit kits and new luminaires. A lookup table was assembled that matched each T12 fixture observed in the sample data to an LED retrofit kit and an LED luminaire based on lumen output.

To determine economic and achievable energy savings potential, we calculated the wattage of each replacement option by dividing the lumen output by the average efficacy (LPW) of DLC-qualified products for each fixture type. We then selected the lowest wattage option (kit or luminaire).

To determine technical energy savings potential, we used the best available efficacy among DLC-qualified products of each fixture type. The lowest wattage option (kit or luminaire) was then selected. Interestingly, we found kits to have higher average and maximum efficacies than equivalent luminaires across all fixture type with one exception, 1x4 troffers.

The DLC qualified products list (QPL) was used to calculate replacement wattages because most commercial lighting programs reference the QPL either as a requirement or recommendation. The QPL can also be queried and downloaded with no limit on data size, which makes it an excellent source for market research.

As noted above, we defined economic savings potential as those T12 replacements that have a two-year payback or less with an average rebate level of 50 percent of incremental cost. Therefore, fixture costs were required for the analysis. On-line distributors such as 1000bulbs.com and warehouse-lighting.com were researched to determine typical costs for each fixture type. These costs did not reflect any volume discounts, which can be significant. We added labor costs assuming an average installation time of 10 minutes per kit and 15 minutes per fixture, at a typical billable rate of $100 per hour for an electrical contractor. 8 No recycling or waste collection costs for T12 components were included in the analysis as this information is not readily available, but would represent additional cost to the business owner. Appendix C provides the fixture lookup table, including average and maximum efficacies and fixture costs.

Once the lookup table was assembled, we calculated the technical, economic, and achievable energy savings potential on a site-by-site basis using the lookup table and operating hours documented for each site. Following the methodology in the TRM, these calculations included an electric HVAC factor to account for interactive effects on the building’s air-conditioning load. A blended rate of $0.10 per kWh, in our experience a typical blended rate for commercial customers in the Midwest, was assumed for the payback calculations.

8 These figures are estimates based on our experience with commercial/industrial lighting projects and were reviewed by a former lighting salesperson for reasonableness.


The site-level savings information was rolled up by region and building size (Small, Medium, and Large) to produce average technical, economic, and achievable energy savings figures. These figures were then multiplied by the business count in each population subset to yield the total savings estimates in Table 14. These figures do not include avoided transmission and distribution losses, which generally vary between 6 percent and 12 percent.

Table 14. Technical, Economical, and Achievable Energy Savings Potential by Region and Size

Category. Results indicate a large technical energy savings potential but more limited economic

savings potential.

Region Total T12

Energy (GWh)

Technical Savings

Potential (GWh)

Achievable Savings Potential (GWh)

Economical Savings Potential (GWh) Size

Metro

Small 71.0 44.3 33.8 0.0

Medium 110.6 68.0 49.5 5.2

Large 259.2 159.2 116.1 0.0

Northeast

Small 45.2 27.8 21.1 0.0

Medium 11.1 7.0 5.9 5.5

Large 43.3 27.3 20.4 0.0

Northwest

Small 10.2 5.9 4.6 0.0

Medium 0.2 0.1 0.1 0.0

Large 12.3 7.9 6.0 0.0

Southeast

Small 51.6 33.6 25.8 0.0

Medium 11.0 6.3 4.9 0.0

Large 38.3 24.9 19.1 4.0

Southwest

Small 35.3 21.7 16.9 0.0

Medium 101.7 68.2 52.8 42.2

Large 15.5 10.3 7.9 5.9

TOTAL (GWh) 816.6 512.4 384.9 62.8

Error! Not a valid bookmark self-reference. summarizes the statewide potential estimates with associated program costs. A 15 percent adder for administration and marketing costs was included in addition to rebate costs; this figure is on the low end of the range of actual non-incentive cost percentages in commercial lighting programs administered by Minnesota customer- and investor-owned utilities as well as alternative CIP providers in 2014 and 2015 (ReportingESP, 2017).


Table 15. Statewide Technical, Economical, and Achievable Savings Potential and Projected Costs.

Conservation Potential GWh Savings

% of T12 Consumption

Total Cost9 (million $) $/kWh10

Technical 512.4 63% N/A N/A

Achievable 384.9 47% $481.5 $1.25

Economic 62.8 8% $11.9 $0.19

At least two important observations can be drawn from these results. First, while the technical energy savings potential is relatively high, only a small subset of T12 upgrades are considered part of the economic savings potential using our definition (a two-year payback requirement with incentives at 50 percent of incremental cost). Second, while approximately 75 percent (384.9/512.4) of the technical energy savings potential could be achieved under a very aggressive scenario of setting rebates at 100 percent of incremental cost, costs would increase exponentially.

Fortunately, the projected cost per kWh saved under the economic scenario, $0.190/kWh, is only slightly higher than the average cost per kWh achieved of all non-residential lighting programs in Minnesota in 2014 and 2015: $0.14/kWh (ReportingESP, 2017)11. This suggests that, at moderate incentive levels, T12 initiatives could be included in most existing programs with little adverse impact on cost-effectiveness.

To understand how the market could respond to changes in incentive level, we modeled a fourth scenario with incentives set to 75 percent of incremental cost while retaining the two-year payback requirement. The results predict that while more savings would be captured, the cost per kWh saved would increase. The impacts of increasing incentive levels are illustrated in Table 16 and Figure 2.

Table 16. Statewide Energy Savings Potential at Varying Incentive Levels.

Average Incentive Level GWh

Savings Total Cost (million $) $/kWh12

50% of incremental cost 62.8 $11.9 $0.19



9 Includes rebate costs plus a 15 percent adder for all other costs associated with delivering a DSM program, including marketing and administration.

10 The cost per kilowatt hour saved presented is for one year of savings only. The lifetime savings cost will be lower by a factor of approximately 15 based on the average lifetime of commercial lighting measures.

11 Ibid.

12 Ibid.


Figure 2. Statewide Energy Savings Potential at Varying Incentive Levels.

The actual dollar amounts used for the 50 percent, 75 percent, and 100 percent incentive scenarios are summarized in Table 17. Note that these amounts factor in labor costs, but do not include any volumetric discounts or recycling costs. The rows corresponding to the most common fixture types are highlighted in yellow. As we discuss later, current rebates for LED retrofit kits and luminaires offered by Minnesota utilities are often lower than these amounts.

To conclude this analysis, we investigated the annual operating hours necessary to achieve a two-year payback or less for each incentive level, again assuming an average blended rate of $0.10/kWh. The results are shown in Table 18, with the most common fixture types highlighted. 13 These results indicate that for 4-foot fixtures, operating hours of 5,500 or greater are required for a two-year payback under the 50 percent scenario. However, if incentives are increased to 75 percent, the hours drop to 2,750 or greater.

The calculated operating hours in Table 18 can be cross-referenced with the table of deemed annual operating hours in Table 10 to understand what building types are likely to have cost-effective T12 opportunities. For example, grocery stores, health services, hospitals, and manufacturing facilities all have deemed operating hours greater than 5,000. It is important to keep in mind that not all customers have a strict two-year payback requirement, however. For instance, government buildings can often accept longer paybacks.

13 For simplicity, this analysis does not include any HVAC interaction. A “*” indicates that it is not possible to two-year payback or less for the given fixture type.


Table 17. Equivalent Dollar Amounts for Modeled Incentive Levels.

Baseline Fixture lumens Proposed Fixture

50% incentive 75% incentive 100% incentive

kit luminaire kit luminaire kit luminaire

T12 U4FT 20W (1) ESMB 1,335 2x2 troffer $49.00 $38.00 $73.00 $56.00 $98.00 $75.00

T12 4FT 34W (1) ESMB 2,225 linear ambient $49.00 $38.00 $73.00 $56.00 $98.00 $75.00


T12 U4FT 20W (2) ESMB 2,670 2x2 troffer $65.00 $75.00 $98.00 $113.00 $131.00 $150.00

T12 4FT 34W (2) ESMB 4,450 20% linear amb/80% 2x4 $65.00 $75.00 $98.00 $113.00 $131.00 $150.00


T12 4FT 60W (1) ESMB 4,673 low-bay luminaire $113.00 $75.00 $169.00 $113.00 $225.00 $150.00

T12 4FT 40W (2) ESMB 4,984 20% linear amb/80% 2x4 $113.00 $75.00 $169.00 $113.00 $225.00 $150.00


T12 4FT 34W (3) ESMB 6,675 2x4 troffer $141.00 $75.00 $211.00 $113.00 $282.00 $150.00


T12 4FT 60W (2) ESMB 8,900 2x4 kit/low-bay luminaire $141.00 $75.00 $211.00 $113.00 $282.00 $150.00








Table 18. Annual Operating Hours Necessary for a Two-Year Payback or Less.

T12 Fixture lumens Proposed Fixture

50% incentive 75% incentive 100% incentive

kit luminaire kit luminaire kit luminaire

T12 U4FT 20W (1) ESMB 1,335 2x2 troffer * 8,600 5,010 4,300 0 0

T12 4FT 34W (1) ESMB 2,225 linear ambient * 8,590 5,750 4,290 0 0

T12 4FT 40W (1) ESMB 2,492 linear ambient 5,980 8,380 2,990 4,190 0 0

T12 U4FT 20W (2) ESMB 2,670 2x2 troffer * * 5,950 7,810 0 0

T12 4FT 34W (2) ESMB 4,450 20% linear amb/80% 2x4 7,650 * 3,830 4,830 0 0

T12 8FT 60W (1) ESMB 4,450 linear ambient * 6,670 5,580 3,340 0 0

T12 4FT 60W (1) ESMB 4,673 low-bay luminaire * * 6,300 4,750 0 0

T12 4FT 40W (2) ESMB 4,984 20% linear amb/80% 2x4 7,550 5,450 3,780 2,730 0 0


T12 4FT 34W (3) ESMB 6,675 2x4 troffer * 6,460 5,430 3,230 0 0

T12 4FT 40W (3) ESMB 7,476 2x4 troffer 8,530 5,010 4,260 2,510 0 0

T12 4FT 60W (2) ESMB 8,900 2x4 kit/low-bay luminaire 5,030 2,900 2,520 1,450 0 0

T12 4FT 34W (4) ESMB 8,900 2x4 troffer 5,540 * 2,770 4,650 0 0


T12 4FT 40W (4) ESMB 10,057 2x4 troffer * * 4,570 7,170 0 0


T12 4FT 60W (3) ESMB 10,324 low-bay luminaire * * * * 0 0

T12 4FT 60W (4) ESMB 13,795 low-bay luminaire 8,580 * 4,290 5,400 0 0


It must also be emphasized that the estimated energy savings potentials represent an early replacement scenario, in which programs target replacement of working T12 fixtures. The fact that the economic energy savings potential is only 8 percent of total consumption does not imply that 92 percent of T12 fixtures will never be replaced; in practice, T12 fixtures will be replaced through what is sometimes referred to as “naturally occurring conservation” as the ballasts die out and businesses find that T12 replacement components are increasingly scarce. A T8 retrofit would be the lowest cost replacement option. However, this would represent a lost opportunity to install a more efficient LED system.


Discussion of Results

Key Conclusions

Several key conclusions can be drawn from the preceding sections:

Despite the best efforts of utilities to eradicate T12s, they still operate in significant quantities: we estimate that T12s account for approximately 10 percent of statewide commercial/industrial lighting demand.

The vast majority of remaining fixtures are 4-foot linear fixtures with two or more lamps.

Small business locations under 5,000 square feet have the highest average T12 power density, 0.284 W/ft2.

T12 fixtures operate 3,210 hours per year on average. Fixtures with 5,500 operating hours or more year are the best retrofit candidates if incentives are equal to roughly 50 percent of installed cost.

The achievable energy savings potential for upgrading T12 fixtures is relatively high, 75 percent of the technical savings potential. However, only a small subset of upgrades qualify as economic savings potential, defined here as having a two-year payback or less with incentives set to 50% of installed cost.

Current incentives for LED retrofit kits and luminaires are insufficient to motivate most commercial customers to change out T12s prior to burn-out, based on energy savings alone.

A word of caution is warranted regarding interpretation of the regional-level results in Table 7, Table 9, Table 13 and Table 14. It must be emphasized that the experimental design of this study was not set up to provide statistically representative results by region; rather, the results are representative of the state as a whole. To provide regionally representative results would require much more extensive data collection. However, the results do serve to illustrate that T12s are found in significant numbers throughout the state and not just in the non-metro areas.

As part of this study, we collected supplemental T12 data from our field staff and key account representatives as they went about their normal business activities, such as performing energy assessments and conducting meetings in customer facilities. This data is summarized in Appendix D. Much of it was provided by our field staff in northwestern Minnesota, and it includes many examples of buildings with high T12 concentrations, belying the relatively low average power density calculated for that region, 0.027 W/ft2 (see Table 12).

It is also worthwhile to point out another geographical limitation of the study: the broad scope of the study, covering the entire state, is not well suited for informing highly localized DSM initiatives. We know from our field work and that of our partners that there are pockets of the state with high concentrations of T12s. One example is the Iron Range in Northeastern Minnesota, which was not selected within the random sample of communities for this study.

It should also be emphasized that we derived the economic and achievable energy savings potentials using a simplistic model of human behavior, in which payback is the sole criteria


driving investment decisions. In reality, most business owners will consider a multitude of criteria when deciding whether to upgrade their lighting. From a purely economic perspective, an organization’s overall economic health may limit its ability to invest in new lighting, even with generous incentives. On the other hand, improved lighting quality and how it relates to productivity, safety and aesthetic appeal may outweigh economic considerations. Finally, less tangible factors, such as public image, for example wanting to appear “green” by installing LED lighting, may be a consideration for some businesses.

Despite these limitations, our results provide insight as to how programs can target remaining T12s. We provide some suggested approaches below and conclude with a discussion of possible TRM updates pertaining to T12s.

Programmatic Approaches

Minnesota utilities have targeted T12 replacements for many years, especially soon before and after the 2012 federal rule. These efforts have usually taken the form of “bounties” or “bonus rebates” paid on top of a standard rebate if the equipment is replacing a T12 fixture. As documented in CIP filings, these efforts were successful in influencing large numbers of customers to change out their T12s.

However, we believe new approaches are needed to root out the remaining T12s. One reason is that the current lighting market is very different from when most of these promotions occurred. At that time, a T8 luminaire or retrofit kit was the standard replacement option for a T12 fixture. Now, there are many more choices available with the advent of LED technologies.

The downside of the having more choices available is the potential for customer confusion and choice overload, especially considering that LED technologies are a moving target. Utilities can help their customers make well-informed decisions by providing non-biased information on the benefits and drawbacks of LED lighting, while at the same time offering financial assistance in the form of incentives or financing.

In the following sections, we describe several key considerations for a successful T12 upgrade program, drawing from the findings in this study and examples of successful programs.

Replacement Product Selection

Only a few years ago, T8 luminaires and retrofit kits were the standard replacement options for T12s. T8 systems are the lowest cost option, and for some customers are a good fit; however, we believe LED systems should be the focus of T12 campaigns for the following reasons:

LEDs are a superior technology, offering higher efficacy, lower lumen depreciation rates, and greater longevity. In addition, LEDs are not adversely affected by on and off cycling, and some products are available with dimming capability.

Labor costs are comparable in T8 conversions versus LED: in either case, an electrician is required.


The Consortium for Energy Efficiency (CEE) no longer maintains a High Performance T8 Qualified Products List, which many programs formerly required for rebate eligibility.

It is arguably a better use of ratepayer funding to incentivize customers once for an LED conversion, compared to potentially twice (once for an initial T8 conversion, then again for an LED conversion at a future date).

Our experience is that most customers today are interested in LEDs rather than fluorescent technologies.

The DLC Qualified Products List of LED products should be referenced either as a requirement for rebate eligibility or as a customer recommendation to help ensure product quality. DLC-certified products must have a lifetime of at least 50,000 hours14 and meet certain performance characteristics including efficacy and color rendering.

Within the broad domain of LED lighting, there are three main product classes that can serve as T12 replacements:

LED Retrofit Kits. Retrofit kits usually consist of a reflector, driver, and one or more LED light strips. The reflector screws into the existing fixture and helps distribute the light. The driver replaces the fluorescent ballast and serves a similar function. The LED strips replace the fluorescent tubes and often attach magnetically to the reflector. Some kits include a replacement lens to provide a refreshed look.

Provided a quality product is selected, kits are often an excellent replacement solution for customers as they offer the same or better performance as a new luminaire, with typically lower upfront material and labor costs. They also don’t suffer from some of the drawbacks of LED tubes described below.

LED Tubes. There are three main types of LED replacement products for fluorescent tubes as classified by Underwriters Laboratories (UL):

o Type A tubes are designed to run off the existing fluorescent ballast and are often referred to as “Plug and Play”.

o Type B tubes contain an internal LED driver and are designed to receive power directly from line voltage, requiring that the existing ballast be bypassed or removed.

o Type C tubes receive power from a remote driver and also require that the existing ballast be bypassed or removed.

o There are also some “combo drive” tubes available that are both ballast-compatible and line voltage-compatible, and thus can function as either a Type A or a Type B tube.

Type A LED tubes are often a great choice for T8 or T5 retrofits because no electrical modifications are required to install them; however, Type A tubes are not compatible with T12 magnetic ballasts. Instead, T12 fluorescent tubes can be replaced with a Type

14 LED lifetimes are generally measured in terms of the L70 rating, meaning the elapsed time for light output to depreciate to 70% of the initial lumen output.


B or C LED tube, both of which require electrical modifications to bypass the fluorescent ballast. Unfortunately, this means that labor costs may be comparable to installing kits or new luminaires. In addition, it is important to keep in mind that LED tubes in general tend to produce light more directionally than fluorescent tubes, which could result in inferior light distribution in some fixtures.

LED Luminaires. If money is no object, replacing the entire luminaire is usually the best solution as the light source, fixture, and optics function as a fully integrated design to optimize ambient light distribution or workplane illumination. However, material and labor costs are usually higher than with retrofit kits or tubes.

Every application is different; because of this, all of these product classes should be considered as eligible replacement options for T12s. However, for targeted T12 replacement campaigns, retrofit kits and certain open strip-type luminaires, which tend to be the least expensive options, may be the most attractive choices. There are also linear high bay fixtures available that can replace existing fluorescent or HID lighting. Table 19 presents a summary of T12 replacement options.

LED Type B Safety Concerns

It is important to highlight some safety concerns associated with LED Type B tubes, which receive power directly from line voltage. The first concern is associated with the socket insulation material: some materials degrade over time when exposed to line voltage, diminishing the electric insulating properties and making them more brittle over time. This concern can be addressed by replacing the existing sockets with new Type B-compatible sockets.

A second safety concern with Type B tubes pertains to shunted versus non-shunted sockets. Shunted sockets have a direct electrical connection between the two pin connectors, sometimes hidden internally.15 If the electrician is unaware of this and rewires the fixture for a non-shunted-compatible LED tube, an electric short will occur at line voltage, potentially causing an electric shock or fire. Fortunately, this is less of a concern for retrofitting T12 fixtures, which use non-shunted sockets.

A final safety concern for Type B tubes is related to mixed lamp types. If a fluorescent tube is mistakenly installed in a fixture with the ballast removed or bypassed, it could overheat and break. This concern can be mitigated by doing a complete lighting retrofit and removing all fluorescent lamps from the premises.

In DSM programs, each of these concerns can be addressed by emphasizing that all electrical modifications should be performed by an experienced, fully licensed and bonded electrician.

15 Shunted sockets are used with T8 and T5 instant start ballasts.


Table 19. Summary of T12 Replacement Options

Product Type Description Advantages Disadvantages

LED Retrofit Kits

Uses existing housing; ballast bypassed and lamps replaced. Some kits include new optics.

- Similar performance to LED luminaires at lower cost

- Avoids safety concerns of Type B tubes

- Use of existing optics changes performance, confounding performance ratings comparisons

LED Tubes

Type A

Lamp receives power from existing fluorescent ballast.

- Not compatible with T12 magnetic ballasts

Type B Includes an integral driver. Lamp receives power directly from line voltage.

- Ease of installation - Slightly more efficient

than Type A

- Safety concerns - Knowledge of shunted

vs. non-shunted sockets required

- More directional light may be a concern

Type C Lamp receives power from a remote driver.

- Ease of installation - Best tube type for

performance and control functionality

- More expensive than Type A and B


Combo Drive

Includes an integral driver. Lamp can receive power directly from line voltage or existing fluorescent ballast.

- Ease of installation - Ideal for facilities with

mixed fixture types

- More expensive than Type A and B


LED Luminaires Replaces fluorescent luminaire.

- Optimal light distribution

- Design flexibility - Higher upfront cost

(varies)

Targeted Outreach

When the 2010 and 2012 fluorescent lighting standards were taking effect and large quantities of T12s were still in operation, programs reaped substantial savings through mass marketing of bonus rebates for T12 change-outs. Much of the low-hanging fruit from these efforts has been harvested, however, pointing to the need for more targeted approaches to reach remaining customers with T12s.

Our study has shown that small business locations tend to have the highest T12 concentrations. Utilities can identify their small business customers relatively easily based on their energy consumption and/or rate code to compile customer lists for direct marketing. We have found, however, that direct marketing alone usually has a very low conversion rate; furthermore, direct marketing does not address the need for customer education regarding LED lighting systems.


A more effective way to reach small business customers is through “boots-on-the-ground.” In rural areas, most small businesses are clustered in small towns. Utility representatives can gather information on T12s by simply going out in the community and performing quick walk-through energy assessments, in which the representatives visually assess building lighting and mechanical equipment to identify energy efficiency opportunities, focusing on the top three to four opportunities in terms of simple payback and savings impact. Usually lighting upgrades are one of the top two opportunities. The auditor can discuss the identified opportunities with the business owner if s/he is present, and follow up with a customized report describing the opportunities, estimated savings and matching incentives from the utility. For administrative efficiency, a report template can be created with predefined savings calculations and language. Direct savings can be achieved through each audit by including a direct install component encompassing measures such as LED light bulbs. This strategy is more difficult to implement in a large city, where small businesses are distributed over a large area, but can still be employed by focusing on individual business districts.

Another approach is to leverage trade allies. Many businesses have longstanding, trusted relationships with local contractors, relying on their expertise when making purchasing decisions (Eureka Recycling, 2014). With regard to lighting, local electrical contractors often fulfill this role. To the extent that a utility can transform these contractors into trade allies through outreach and education, they can function as a highly effective adjunct sales force for the utility. Minnesota Power successfully used this approach, as described in the following section.

Incentives, Buydowns, and Copays

While mass-marketed bonus offers alone are unlikely to be effective in influencing remaining customers with T12s, our findings suggest that incentives are still critical for lowering the upfront cost and reducing the payback period. Most existing rebates for LED products appear to be insufficient to achieve our benchmark of a two-year payback in most cases. Therefore, utilities should consider offering a bonus rebate on top of their standard rebate for T12 upgrades. Table 17 can be used as a guide for setting a specific amount.

As described previously, we estimate that achieving the economic savings potential of 62.8 GWh would cost $0.19 per kWh saved. This cost level is higher than the statewide average cost for non-residential lighting programs of $0.14 per kWh, but within the range of existing programs. Given that our data tells us that T12s are most prevalent in small businesses, a T12 initiative can be viewed from the standpoint of ratepayer equity: higher incentives would provide a means for this typically underserved customer segment to participate and benefit from CIP.

Some utilities may also find success with a midstream incentive model, in which contractors are incentivized to promote a specified product type. Minnesota Power recently implemented a successful T12 campaign in International Falls by offering a $10 per fixture “spiff” (an immediate incentive payment or bonus for the sale of a specific product) to local electrical contractors for upgrading T12s to eligible LED luminaires or kits. This campaign came about as a result of information gathered through energy assessments of local businesses, which


identified high quantities of T12s. According to utility staff, the program was successful in influencing a sizable number of customers to switch to LED technology (Gallagher, 2017).

A third way that utilities could provide financial assistance is through a buydown program. Under this model, the utility purchases an inventory of replacement fixtures on behalf of a group of customers to obtain wholesale pricing. This initiative allows small business customers to obtain attractive pricing on LED products that larger customers can secure through volumetric discounts. If necessary, a customer copay can be charged to improve program cost-effectiveness. Though this model has not been tested in the commercial sector for linear LED products to our knowledge, experience in the residential sector with marginally cost-effective products such as smart thermostats and LED bulbs has shown that, with appropriate marketing and messaging, customers find a copay offer with direct install attractive (Heffron and Carroll, 2016).

Messaging

Effective messaging is essential for selling energy efficiency. Although the lower operating costs of LED lighting are compelling, program staff should not overlook the non-energy benefits associated with LED lighting, including better color rendering, instant on and off, dimming capability (in some products), and lower lumen depreciation rates. The long lifespan of LED lighting (50,000 hours for DLC-rated products compared to 24,000 for High Performance T8 Lamps (DLC, 2017 and CEE, 2009) may also be compelling for some customers in terms of reduced maintenance costs, especially in high bay applications. Depending on the specific application, these characteristics may be equally or more important than lower energy costs.

For lack of a better term, “scare tactics” can also be used; namely, staff can point out that the T12 ballasts are likely nearing the end of their life and that new T12 lamps and ballasts are becoming increasingly rare and expensive.

TRM Recommendations

One of the objectives of this study was to review treatment of T12 replacements in the Minnesota TRM and recommend possible changes or updates in light of the study findings. Currently, there are two T12 retrofit measures in the TRM falling under the C/I Lighting End Use category: T12 Up to 4-Foot Retrofit and T12 8-Foot Retrofit. Each measure is comprised of multiple baseline and proposed fixture pairings delineated in TRM Appendix B, a standalone Excel spreadsheet. Each pairing includes baseline and proposed fixture descriptions, baseline and proposed wattages, and the proposed fixture cost.

The proposed fixtures in the two T12 measure groups currently include only T8s and T5s. If Commerce wishes to continue T12 rebates in 2018, additional pairings for T12 to LED conversions could be added. The current deemed costs for T8 and T5 fixtures appear to be reasonably accurate.

An important parameter for any TRM measure is the lifetime. The lifetime typically reflects the average lifespan of the proposed equipment, considering both the physical lifetime of the equipment and how long it is typically operated in practice before being replaced.


Minnesota TRM Version 2.0 includes a deemed lifetime of 1 year at full savings for each T12 measure in 2017. This lifetime does not reflect the lifetime of the replacement equipment, but instead reflects the estimated remaining useful life of T12 ballasts. The lifetime was originally set to four years in 2013 based on DOE supporting analysis for the 2000 and 2011 ballast rules and decreased annually to one year in 2016. The one year lifetime was extended through 2017 in TRM 2.0 as a result of this study.

The actual remaining useful life of operable T12 fixtures is difficult to estimate. Within a certain product class, actual lifetimes will vary even if operating conditions are identical because of manufacturing variations. Moreover, in practice, operating hours vary greatly: ballasts with low operating hours could continue to operate for many years, while fixtures with high operating hours may have very little life left. Operating conditions also vary from site to site: for example, fixtures with more on/off cycling would be expected to have lower remaining useful life. Given all this uncertainty, continuing the one-year measure life through 2018 seems appropriate.

Adding LED kits and luminaires as replacement options for T12 fixtures would then allow utilities to claim the full savings for an LED conversion over a one-year measure life. This is a conservative approach because the LED will produce savings over its entire lifespan, which we estimate at 15.6 years for a DLC-certified fixture16, but with the baseline changing to a standard T8 in year two, when the T12 ballast is assumed to fail. This approach would ignore all savings occurring in years 2-15, but still allow the utility to claim the full savings from upgrading to LED in year one. An example conversion from a T12 2x4 troffer with two 34W lamps (fixture wattage = 67W) to a 20W LED retrofit kit is illustrated below in Figure 3 and Table 20. .

Figure 3. Lifetime Savings Example for LED Kit.

16 Based on a 50,000 hour lifetime divided by average operating hours of 3,210 hours per year measured in this study.


Table 20. Tabulated Representation of Lifetime Savings Example for LED Kit.

Year Baseline Baseline kWh Proposed kWh Claimed Savings

1 Existing T12 2x4 troffer 215.1 64.2 150.9

2 T8 2x4 troffer 179.8 64.2 0

3 T8 2x4 troffer 179.8 64.2 0

4 T8 2x4 troffer 179.8 64.2 0

5 T8 2x4 troffer 179.8 64.2 0

6 T8 2x4 troffer 179.8 64.2 0

7 T8 2x4 troffer 179.8 64.2 0

8 T8 2x4 troffer 179.8 64.2 0

9 T8 2x4 troffer 179.8 64.2 0

10 T8 2x4 troffer 179.8 64.2 0

11 T8 2x4 troffer 179.8 64.2 0

12 T8 2x4 troffer 179.8 64.2 0

13 T8 2x4 troffer 179.8 64.2 0

14 T8 2x4 troffer 179.8 64.2 0

15 T8 2x4 troffer 179.8 64.2 0


Conclusion

This study has developed a statistically rigorous baseline estimate of the energy consumed by remaining T12 fixtures in Minnesota. In doing so, it has confirmed what many observers suspected, that this outdated technology lives on.

From the field data collected, we know that T12s are most prevalent in small business locations under 5,000 square feet. They are also found throughout the state, though pockets of high concentrations are known to exist as a result of utility outreach activities. Utilities can effectively pinpoint areas of T12s as well as gather valuable information on other efficiency opportunities through small business “blitzes” and other forms of targeted outreach.

While many of these fixtures could be upgraded as they fail through “natural conservation,” utilities have the opportunity to proactively reach out to customers to help them convert directly to LED fixtures rather than less costly and less efficient T8s. Outreach can take the form of energy assessments offered by utility representatives coupled with T12 bounty offers, or a midstream program leveraging trade allies to identify and promote upgrades. Bonus incentives on top of existing LED rebates are likely needed to motivate customers with working fixtures to participate in T12 to LED upgrades.

The explosion of new LED products has benefited consumers by offering more options and a superior technology. However, the potential for buyer confusion and choice overload has grown just as fast. Energy efficiency programs can help educate customers to make an informed decision when upgrading lighting by providing unbiased, expert advice in addition to offering financial incentives for quality products.


References

Appliance Standards Awareness Project (ASAP). “General Service Fluorescent Lamps.” (http://www.appliance-standards.org/node/6802), accessed January 12, 2017.

Appliance Standards Awareness Project (ASAP). “Fluorescent Lamp Ballasts.” (http://www.appliance-standards.org/node/6811), accessed January 12, 2017.

Consortium for Energy Efficiency. “CEE High-Performance T8 Specification.” Revised June 30, 2009.

Department of Energy (DOE). 2010. “Fluorescent Lamp Ballasts Preliminary Analytical Tools: National Impact Analysis.”

Department of Energy (DOE). 2000. “Fluorescent Lamp Ballast Technical Support Document for the Final Rule, 2000.”

Design Lights Consortium. 2016. “Technical Requirements V4.1.” (https://www.designlights.org/solid-state-lighting/qualification-requirements/technical-requirements/), accessed February 1, 2017.

Energy Information Administration (EIA). 2016. Form EIA-861 data files. (https://www.eia.gov/electricity/data/eia861/), accessed December 29, 2016.

Environmental Protection Agency (EPA), Greenhouse Gas Equivalencies Calculator. (https://www.epa.gov/energy/greenhouse-gas-equivalencies-calculator), accessed February 27, 2017.

Eureka Recycling. 2014. “Energy Conservation in the Food Service Sector.” Conservation Applied Research & Development (CARD) Program Final Report. Prepared for Minnesota Department of Commerce, Division of Energy Resources.

Gallagher, Tim. Minnesota Power. February 27, 2017 phone conversation.

Heffron, Jim and Carroll, Ed. 2016. “Driving Adoption of Marginally Cost Effective Measures through Customer Copay.” Proceedings of 2016 ACEEE Summer Study on Energy Efficiency in Buildings.

Minnesota Rules. 2015. Chapter 1323, Commercial Energy Code. Table C405.5.2(1), Interior Lighting Power Allowances: Building Area Method.

National Lighting Product Information Program (NLPIP). “Cathode-Disconnect Ballasts.” Specifier Reports Volume 2, Number 1. June 1993.

ReportingESP. (www.energyplatforms.com), accessed February 14, 2017.

State of Minnesota Technical Reference Manual (TRM) for Energy Conservation Improvement Programs. Version 1.2.

State of Minnesota Technical Reference Manual (TRM) for Energy Conservation Improvement Programs. Version 2.0, April 2016.

http://www.appliance-standards.org/node/6802

http://www.appliance-standards.org/node/6811

https://www.designlights.org/solid-state-lighting/qualification-requirements/technical-requirements/

https://www.eia.gov/electricity/data/eia861/

https://www.epa.gov/energy/greenhouse-gas-equivalencies-calculator

http://www.energyplatforms.com/


World Health Organization (WHO). “Steps in Applying Probability Proportional to Size (PPS) and Calculating Basic Probability Weights.” (http://www.who.int/tb/advisory_bodies/impact_measurement_taskforce/meetings/prevalence_survey/psws_probability_prop_size_bierrenbach.pdf), accessed April 1, 2016.

http://www.who.int/tb/advisory_bodies/impact_measurement_taskforce/meetings/prevalence_survey/psws_probability_prop_size_bierrenbach.pdf

http://www.who.int/tb/advisory_bodies/impact_measurement_taskforce/meetings/prevalence_survey/psws_probability_prop_size_bierrenbach.pdf

Appendix A


Appendix A. Data Collection Template

Appendix B


Appendix B. Derivation of Statewide Lighting

Load

1. According to the EIA Commercial Building Energy Consumption Survey (CBECS), the average energy use intensity (EUI) for commercial lighting in the West North Central region was 2.3 kWh per square foot in 2012 (the most recent year available at the time of this report). The overall EUI for commercial buildings was 12.5 kWh/ft2.

2. Total commercial and industrial (C/I) energy use in Minnesota in 2015 was 44,841,055 kWh according to the EIA-861 database.

3. The total square footage of C/I buildings in Minnesota equals 44,841,055 kWh / (12.5 kWh/ft2) = 3,587,284,400 ft2.

4. Total C/I lighting energy in Minnesota is estimated to equal 2.3 kWh/ft2 * 3,587,284,400 ft2 = 8,250,754,120 kWh.

5. Dividing 8,250,754,120 kWh by 3,226 hours, the weighted-average operating hours measured from the sample data, yields 2,557,580 kW (2,558 MW), the total estimated C/I lighting load in Minnesota.

Appendix C


Appendix C. Fixture Lookup Table

Baseline T12 Fixture* Baseline

Watts

DLC Standard LED Retrofit Kits DLC Standard LED

Luminaires Best

Available

Proposed Watts

Mat'l Cost Labor

Total Cost

Proposed Watts

Mat'l Cost Labor

Total Cost

Proposed Watts

T12 4FT 34W (1) ESMB 42 17.7 $80.95 $16.67 $97.62 20.2 $50.00 $25.00 $75.00 13.7

T12 4FT 40W (1) ESMB 41 19.8 $80.95 $16.67 $97.62 19.2 $50.00 $25.00 $75.00 15.4

T12 4FT 60W (1) ESMB 81.9 39.1 $85.53 $16.67 $102.20 35.9 $129.00 $25.00 $154.00 24.3

T12 4FT 34W (2) ESMB 67 39.6 $114.00 $16.67 $130.67 43.0 $125.00 $25.00 $150.00 27.1

T12 4FT 40W (2) ESMB 87 44.3 $114.00 $16.67 $130.67 48.2 $125.00 $25.00 $150.00 30.4

T12 4FT 60W (2) ESMB 142.2 79.1 $265.00 $16.67 $281.67 86.0 $125.00 $25.00 $150.00 54.3

T12 4FT 34W (3) ESMB 104 59.3 $208.50 $16.67 $225.17 64.5 $125.00 $25.00 $150.00 40.7

T12 4FT 40W (3) ESMB 141 66.5 $208.50 $16.67 $225.17 72.2 $125.00 $25.00 $150.00 45.6

T12 4FT 60W (3) ESMB 203.6 91.8 $265.00 $16.67 $281.67 99.7 $125.00 $25.00 $150.00 63.0

T12 4FT 34W (4) ESMB 144 79.1 $265.00 $16.67 $281.67 86.0 $125.00 $25.00 $150.00 54.3

T12 4FT 40W (4) ESMB 172 89.4 $265.00 $16.67 $281.67 97.2 $125.00 $25.00 $150.00 61.4

T12 4FT 60W (4) ESMB 262.6 122.7 $265.00 $16.67 $281.67 133.3 $125.00 $25.00 $150.00 84.2

T12 8FT 60W (1) ESMB 74 35.3 $69.00 $16.67 $85.67 40.4 $100.00 $25.00 $125.00 27.4

T12 8FT 75W (1) ESMB 94 39.6 $69.00 $16.67 $85.67 45.2 $100.00 $25.00 $125.00 30.7

T12 8FT 60W (2) ESMB 113 70.7 $138.00 $16.67 $154.67 80.8 $160.00 $25.00 $185.00 54.8

T12 8FT 75W (2) ESMB 145 79.8 $138.00 $16.67 $154.67 91.3 $160.00 $25.00 $185.00 62.0

T12 U4FT 20W (1) ESMB 21 12.2 $81.67 $16.67 $98.34 13.4 $88.00 $25.00 $113.00 8.5

T12 U4FT 20W (2) ESMB 53 24.3 $81.67 $16.67 $98.34 26.8 $88.00 $25.00 $113.00 16.9

*Fixture Code: T12 <Lamp Length> <Lamp Wattage> <No. of Lamps> <Ballast Type>

Appendix D


Appendix D. Supplemental T12 Data

The following table summarizes additional data collected through normal business activities outside of the sample data.

Region

No. of Sites T12 Load (kW) T12 Energy (kWh) T12 Avg. Watts

per Sq Ft Size

METRO

Small 1 1.3 3,042 0.520

NW

Small 4 20.3 2,124,538 8.111

Medium 3 17.0 50,668 2.261

Large 5 19.0 182,078 0.581

SE

Small 4 16.3 30,885 6.510

Medium 1 2.1 6,321 0.277

Large 1 19.1 89,245 0.583

SW

Small 1 0.0 0 0.000

Medium 2 20.4 59,473 0.624

Turning Off T12 Lighting… For Good!mn.gov/commerce-stat/pdfs/card-t12-lighting.pdf · vapor in a...

Documents

Transcript of Turning Off T12 Lighting… For Good!mn.gov/commerce-stat/pdfs/card-t12-lighting.pdf · vapor in a...