The US economy is only about 33% energy efficient today, with two-thirds of primary energy production lost before it’s put to use. Energy efficiency improvements will be a huge part of sustaining economic growth moving forward.
What is the total energy efficiency of the US economy? More precisely, what percentage of primary energy input, whether from fossil fuels or renewables, is actually used for our benefit? The Lawrence Livermore National Laboratory (LLNL) studies energy flows within the US economy, and produces the diagram above annually. For 2008, LLNL calculates that the nation’s energy efficiency is roughly 42%, with the remaining 58% of energy lost primarily as wasted combustion heat .
By itself, having a 42% efficient economy indicates that significant improvements in our energy security are possible through efficiency improvements alone. But there is ample evidence that the US is significantly less efficient than the diagram indicates. One key assumption in particular biases the calculation: LLNL assumes that residential, commercial, and industrial users of energy are 80% efficient in their end use of energy. The primary uses of energy in American homes are heating, cooling, and water heating, and these activities are closer to 50% efficient than 80% efficient in the average American home . Even the most efficient lighting is only about 20% efficient, and huge number of computers and gadgets using and powering the internet can be incredibly inefficient . While industrial users of energy have an economic incentive to prevent waste, they can’t avoid the inefficiencies inherent in lighting and mechanical engines. Finally, transportation in the US is probably closer to 20% efficient, rather than 25% as used in the LLNL study .
In place of the 80% efficiency used by LLNL, we can substitute a 50% efficiency estimate for residential and commercial users, a 75% estimate for industry, and a 20% estimate for transportation. Using these estimates of end-use efficiency, total energy efficiency in the US economy is around 33%. The US has huge room for improvement, which should provide hope for our energy future. If the US were able to improve its energy efficiency from 33% to 50%, primary energy usage could be cut by a third. The US could supply today’s annual energy needs with 67 quads (quadrillion BTU) instead of 102 quads. That’s $650 Billion worth of energy at today’s prices !
During the energy shock of the late 70’s and early 80’s, the US decreased the energy intensity of its economy by as much 5% per year. Given the dual constraints of rising world energy demand and Peak Oil, a similar effort may be required soon. It’s good to know that the US has plenty of room to improve, and to know that improvements in efficiency can generate huge savings.
 Neither the economy nor even a single power plant can ever by 100% energy efficient. The heat engines (internal combustion engines, fossil-fuel power plants, etc) used to run most of modern society cannot exceed certain theoretical limits governed by the laws of thermodynamics, so that even the most efficient combined-cycle power plants today are only about 60% efficient.
 American residential energy use is dominated by hvac usage and water heating, as the EIA shows in table 14 of its energy usage survey. Household furnaces range from 65% to 90% in efficiency, but the typical house loses 25% of its heat through its windows alone, so that heating a house is perhaps 50% efficient overall. Gas water heaters are 50-70% efficient. None of these numbers take into account time when a home is unoccupied, when energy efficiency is effectively 0% if systems are not turned off.
 Incandescent light bulbs aren’t even 5% efficient, while fluorescent bulbs reach around 15% efficiency. During the summer and in warmer parts of the country, the majority of the energy used in lighting (lost as heat) must be counteracted with air conditioning! Computers and data centers are also incredibly inefficient, particularly when measured from a standpoint of average cpu utilization. Servers in data centers are typically configured with enough cpu power to sustain peak activities like handling the Christmas rush at an ecommerce website. As a result, they spend most of their time at very low average utilization, and probably run at less than 10% energy efficiency.
 Internal combustion engines are only about 20% efficient in real-world driving conditions, and this may not take into account extraordinary efficiency losses caused by traffic jams, which themselves result in billions of dollars worth of economic losses annually.
 The EIA Kids’ page provides convenient numbers on the BTU content of different fuels (if only adults knew this much!). We can calculate the rough average price of a BTU by averaging the cost per BTU of oil and the cost per BTU of coal. According to the EIA, a ton of coal costs roughly $50, and contains 19.98 million BTU, for a cost of $2.50 per million BTU. A barrel of oil costs roughly $80 and contains 5.8 million BTU, for a cost of $13.80 per million BTU. A simple average gives us a price of roughly $8 per million BTU. Saving 35 quadrillion BTU of energy with changes at power plants thus equates to a savings of $280 Billion per year. If the energy efficiency changes occur at the point of end use, however, the savings could be much greater, since higher value forms of energy like electricity are much more expensive. One kilowatt-hour of electricity equates to 3412 BTU, and costs roughly 10 cents, which equals $29.30 per million BTU. If the 35 quadrillion BTU of energy efficiency savings all occur at point of end use, approximately $1 trillion in annual savings are possible! In reality, efficiency gains will occur across the system, so an average of the two estimates yields a savings estimate of $650 Billion per year.