Clark Silcox, NEMA General Counsel

In the literature pertaining to energy conservation and Standards and codes, there is a concept known as Naturally-Occurring Market Adoption (NOMAD).  This is defined to mean the proportion of energy savings that would have taken place in the market even if energy conservation Standards had not been adopted, and it represents an examination of trends based on market data.   The NOMAD experience is highly relevant to what has already transpired and is continuing to transpire with respect to light bulbs, and it is central to NEMA position on light bulb regulation:  to wit, the market has already done the job that mandatory regulatory Standards will do.  The difference between a regulatory and non-regulatory outcome is, we submit, miniscule and in a few years will be negligible approaching zero without regulation.  The speculation of some that consumers will change their behavior and start buying more incandescent bulbs in the absence of additional regulation enjoys no empirical support and is contrary to the weight of data documenting a steady sixteen-year decline in incandescent lamp shipments.

As explained below, the estimated additional cost of electricity to the U.S. household because general service incandescent lamps are not eliminated from the market by mandatory regulation is equivalent, on average, to two brand label chocolate bars per year in 2021 and fewer than two chocolate bars per year in 2023.  And the need for more electric power plants in the absence of regulation of general service light bulbs?  None.  The net difference in domestic utility-scale electricity generation between a regulatory and non-regulatory scenario for general service light bulbs is just over 1/10^th of one percent in 2021 and less than 1/10^th of one percent in 2023.

The trends based on market data are displayed in the three charts below.  Domestic U.S. shipments of general service incandescent lamps have been declining since 2003, long before any regulatory Standards took effect.  At that point, shipments of more efficient compact fluorescent lamps with a longer socket life were just starting to show significant growth.  That growth accelerated in 2007 and continued robustly through 2014, thereafter shipments of general service LED lamps started to accelerate in 2015.

The following chart compares the shipments of each of the three technologies of general service lamps since 2012 and includes a forecast through 2023.  Incandescent and CFL bulbs continue their descent, and owing to the dominant presence of longer-life LED bulbs in sockets, even LED shipments start to decline as the demand for replacement bulbs at any point in time shrinks.

Shipments of light bulbs, however, are not a direct measure of electricity consumption; it is their presence in lamp sockets that is important to the measurement of electricity use.  Shipments are one input to the assessment of how many incandescent, CFL, or LED bulbs are in sockets; another important input is the service life of the bulb --- how long it resides in a socket.  Longer life CFLs and LED crowd out the ability the shorter-life incandescent bulb to substitute in a socket at any point in time and therefore influences what type of bulb remains in lamp sockets.  Ultimately shipment data combined with assumptions about light bulb service life and data-derived light bulb wattage enables reasonable estimates of electricity use by these lamps over time.  Here are the estimates of the installed base (or stock) showing general service light bulb socket penetration by technology that we have calculated for 2012, 2019, 2021, and 2023 in the absence of regulation.

Separately, we examined the impact on lamp socket penetration of an energy conservation Standard that could not be met by a general service incandescent lamp.  Incandescent light bulbs do not disappear from sockets overnight, but they do disappear from store shelves sooner and over time from sockets.  Here is the estimate of socket penetration by the end of 2021 and 2023 under a scenario that assumes that general service incandescent lamps are banned at some point in 2020. The model results in approximately 5 percent fewer (5 percent vs. 10 percent) general service incandescent lamps in sockets in 2021 under a regulatory scenario compared to the baseline scenario; by 2023 there are 4 percent fewer (2 percent vs. 6 percent) general service incandescent lamps in sockets under a regulatory scenario compared to the baseline scenario.

There is a consensus estimate that there are approximately 3 billion general service lamp sockets in the United States.  Our estimate of total sockets, derived from shipment data and estimates of lamp service life, consistently supports the consensus estimate.  The exact number is not important with respect to making comparisons of electricity use here.  With data-supported estimates of the various wattages of incandescent, CFL, and general service LED light bulbs in sockets, an estimate of electricity use by general service lamps can be made considering the lamp stock allocations displayed above.  Displayed below are our estimates of general service lamp electricity use in 2012, 2019, and forecast for 2021 and 2023.  From 2012 to 2019 electricity use by general service lamps has fallen by just over fifty percent.  Some of this decline is attributable to the shift from traditional incandescent lamps to halogen incandescent lamps that use 28 percent less electricity as required by federal energy Standards. However, most of the decline is due to the increasing socket occupancy by the CFL and general service LED bulbs that is not driven by any mandatory energy Standards.  The non-regulatory forecast projects that electricity use by general service lamps will decline 22 percent from 2019 to 2021, primarily due to continuing replacement of CFLs and incandescent bulbs by general service LEDs and an additional 14 percent from 2021 to 2023 for the same reason.

The graph above shows the comparative outcome for electricity use by general service lamps with (blue line) and without (red line) regulation indicating that the two scenarios begin to converge by 2023.  The graph confirms that most of the energy savings from the adoption of energy savings has already occurred.  This difference over the baseline scenario results in our energy savings estimate of 5.08 billion kilowatt hours by the end of 2021 or 0.116 percent of 4.28 trillion kWh utility-scale electricity generation in 2021.  5.08 billion kilowatts hours represent .017 quads of energy, which is about 17/1000th of a percent (0.01683 percent) of the 101 quads of energy consumed in the United States in 2018.  Figure 9 projects that the difference between the baseline scenario (red line) and the ban scenario (blue line) and shows the difference narrowing by the end of 2023 as the electricity use under the two scenarios begin to converge with the net difference declining to 4.0 billion kWh difference between the two scenarios, less than one-tenth of one percent (0.094 percent) of annual utility-scale electricity generation in 2023 as the baseline scenario and the ban scenario converge.

To appreciate the electricity consumption pattern by general service lamps over the 2012 – 2023 timeframe, the table below shows the percentage of electricity use by general service lamps as a percent of utility-scale electricity generation.

As this table demonstrates, electricity use by general service lamps (as a percent of U.S. utility-scale electricity generation) declines by over half from 2.46 percent to 1.19 percent from 2012 to 2019.  The historical shipment data provided at the outset of this discussion provides the explanatory context for this development.   The non-regulatory baseline scenario forecasts an additional decline of 0.28 percent, while the regulatory scenario would forecast a decline of 0.4 percent by 2021.  The “net” energy savings after accounting for naturally occurring market adoption is just over one-tenth of one percent of domestic utility-scale electricity generation in 2021 and less than one-tenth of one percent of domestic utility-scale generation in 2023.  As the regulatory ban scenario sees only a small decline in electricity use after 2021 and the baseline scenario begins to converge toward the ban scenario over next few years, the incremental difference in energy savings year over year from a ban scenario is not meaningful in this context and is too small to have an impact on domestic power plant requirements.  Under both scenarios, electricity use from general service lamps continues to decline, just as it has been for more than a decade now.

To put this into the perspective of the household, the U.S. Census estimates that in 2018 there were 127.59 million households in the United States.   Dividing the estimated 5.08 billion kilowatt hours of net energy savings in 2021 by 127.59 million households yields an average of 39.81 kilowatt hours of electricity per household in the U.S. in 2021. According to the most recent information from the U.S. Energy Information Administration (EIA), the average electricity usage per household in the United States in 2018 was 10,972 kilowatt hours.  39.81 kilowatt hours represent approximately 0.0036 percent of the average household electricity use in the United States.  The average price of electricity in the United States in 2019, according to EIA, is $0.1329 per kilowatt hour putting the incremental average household cost of electricity in the baseline scenario over the regulatory ban scenario in 2021 at $5.29 per year or $0.44 per month due to the continuing but dwindling presence of general service incandescent halogen lamps in household lamp sockets.

For 2023, the differential impact between the two scenarios is even smaller.  Using the same calculations that were just made for 2021, and dividing the estimated 4.0 billion kWh of net energy savings in 2023 by 127.59 million U.S. households (from 2018 estimates) yields an average of 31.35 kilowatt hours of electricity per household in the U.S.  per year attributable to the continuing presence of general service incandescent lamps in sockets under the baseline scenario versus the regulatory ban scenario.  That is 0.003 percent of the average annual household electricity use.  At $0.1329 per kilowatt hour, 31.35 kilowatt hours of electricity represents an average household cost of $4.16 per year ($0.34 per month) due to the continuing but dwindling presence of general service incandescent lamps in household lamp sockets.  With private label chocolate candy currently costing about $2.75 per piece, the difference in annual household electricity cost between the two scenarios is less than the cost of two pieces of chocolate candy per year.

It is for this reason, NEMA has encouraged the Department of Energy to apply one of the tenets of the Clinton Administration’s 1996 Appliance Efficiency Rulemaking Process Rule and consider the efficacy of non-regulatory alternatives such as the continuation of the Energy Star program for general service lamps in lieu of mandatory regulation.   The history and market projections for the general service lamp class of covered products present a unique set of circumstances where the market is already strongly aligned with the nation’s energy conservation policy as well as environmental policies reducing power plant emissions.

The NOMAD phenomenon is occurring beyond the general service lamp category.  The same trend is occurring in reflector lamps, decorative lamps, and tubular lamps primarily used in commercial operations (long dominated by fluorescent tubular lamps).  Incandescent lamp shipments are declining in virtually every category.  Market adoption of LED versions of these lamps is now accelerating.

Applying the same general socket penetration analysis for reflector lamps that we applied above, CFL and LED are estimated to occupy 55 percent of large diameter reflector lamp sockets, and incandescent reflector lamps occupy an estimated 45 percent of those sockets by the end of 2019.  Large-diameter reflector lamp sockets are a fraction (~17-18 percent) of general service lamp sockets in 2019.  LED reflector lamp shipments exceeded incandescent reflector lamp shipments for the first time in 2018 and market adoption of LED reflector lamps is accelerating.  Year-to-date 2019 import data and NEMA domestic shipment data show that the LED reflector lamp is now dominating this category with LED reflectors representing an estimated two-thirds of large diameter reflector lamp shipments. We estimate that by 2023, based on current trends, LED reflector lamps will occupy an estimated 88 percent of large diameter reflector lamp sockets under a naturally occurring market adoption scenario.

These examples illustrate just two components of the larger lighting efficiency success story (household, commercial, and industrial) on which we have previously reported.  Without a doubt these conclusions demonstrate that lighting has made the largest contribution to reducing electricity use in this century compared to all other equipment and appliances that use electricity.  It underscores what the lighting industry and NEMA  have been explaining all along; whether the government regulates or more realistically allows NOMAD to continue apace, energy conservation in lighting is not “rolling back” but leading forward as the marquis example of industry and consumers getting the job done.