GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted

Community and Q&A

Centralized or decentralized fossil fuel use, which is most efficient?

Jonathan Blaney | Posted in Energy Efficiency and Durability on

Is burning gas/coal/oil in centralized electric generation plants better for the environment than burning them in a residence or business for heat/hot water.  For example, is a heat pump better than your gas or oil boiler?  Are there studies?

GBA Prime

Join the leading community of building science experts

Become a GBA Prime member and get instant access to the latest developments in green building, research, and reports from the field.


  1. John Clark | | #1

    Physics tells you that the closer you are to the energy source the potential you have for greater efficiency. This issue is at what price. The cost of that generation varies (raw materials, environmental via regulatory requirements, govt policies favoring one energy source over another) etc.

    Take note of the amount of rejected energy per source.

    1. Expert Member
      Dana Dorsett | | #2

      Physics also tells you that generating electricity has much lower raw thermal efficiency than burning it simply for heat, independent of the transmission losses.

      Burning natural gas in a best-in-class combined cycle power generator at it's optimal firing rate is at best 60% thermally efficient right at the primary winding of the first transformer, before the power leaves the plant. Burning natural gas in a 95% AFUE condensing boiler already has it beat by 50%.

      But even after transmission losses bringing down the combined cycle plant's net efficiency to 50% when delivered to the load, a heat pump operating even at a paltry COP of 2 brings it up to 100% efficiency when deliver to the desired end product (heating BTUs entering the house), thus beating the condensing gas boiler. At an HSPF of 10 (=seasonal average COP of 2.93) the heat pump beats the condensing gas burner hands-down, if the primary fossil-burner on that grid is a combined cycle gas unit.

      Typical single cycle gas or oil fired peakers run about net 30-35% efficiency, as do the best sub-critical coal fired base load plants. (Supercritical coal can hit the mid-40s, but that doesn't represent the existing coal fired fleet anywhere.) At 30% net efficiency and an HSPF of 10 you're still looking at about 88% total efficiency from fuel to BTUs inside the house.

      Bear in mind that a grossly oversized oil or cast iron gas boiler won't hit it's AFUE numbers either. AFUE testing is done at a duty cycle that presumes a 1.7x oversize factor for the load at the 99% outside design temp. At 3x oversizing for the 99% load one would have to knock 8-15 % off the nameplate AFUE or steady-state efficiency numbers, unless smart heat purging boiler controls are being used. (Many/most new boilers have these types of controls, but not all, and very few boilers 10+ years old would have them.) See Table 3, comparing the oversizing efficiency hit relative to steady state for system #3 with the heat purging controller vs. the rest:

      To find out your as-used oversize factor, run this analysis, and compare the implied heat load to the DOE output numbers of the boiler:

  2. Joel Cheely | | #3

    Also factor in: the leakage rate of natural gas before it gets to your home (reported by some to be as much as 3%), the first cost of two utilities being run, and the cost to maintain two utilities to your home.
    Now if we all had gas fired microturbines with heat and cooling trigeneration we could eliminate grid electric and just have natural gas.

  3. Expert Member
    Zephyr7 | | #4

    Natural gas used directly for heating will generally be more efficient in terms of delivered BTUs or heat to the home per unit of natural gas consumed. The reason is that electrical generation is not 100% efficient, and there are additional losses in transmission lines delivering the electricity from the power plant to your home. More energy conversions generally means less overall system efficiency.

    Natural gas transmission losses and other incidentals are there regardless of if the gas is delivered to a power plant or to your home, so that part doesn’t really “count” here.

    Now if you were using the natural gas to generate electricity, it’s more efficient to use power from the grid. The reason is because the entire point of having a large-scale power grid is to improve overall system efficiency: the large grid allows the most efficient plants to run the most, and it also allows for those plants to operated near full capacity at all times for maximum efficiency. Typical home generators will have a much lower average utilization as a percentage (ie you need a 10kw unit for peak loads but rarely actually draw more than 1kw from it), so they have much lower operational efficiency.

    Dana is correct that using a heat pump may get you more useable heating BTUs compared with burning natural gas directly for heating, but I don’t think of that as efficiency of gas use as much as it’s a different type of heating system that uses less input energy to accomplish the same amount of heating. semantics in a way, but an important distinction in my opinion. Using natural gas to run a home generator that then runs your heat pump is probably going to be less efficient than using grid electricity to run your heat pump, for example, even if the grid was being supplied by natural gas power plants.

    I very much doubt natural gas system “leakage” is anywhere near 3%. My guess is actual system leakage is what is known as a “trace amount”, an almost immeasurably small percentage. 3% of the natural gas used on even a daily basis would be an absolutely massive amount of methane being leaked out and would cause all kinds of problems. My guess is that “3%” number would include gas used to power transmission equipment (compressor stations, pumps along pipelines that keep the pressure up) and the like, which is much more reasonable than the assumption that that gas is just vented to the atmosphere.


  4. Expert Member
    Dana Dorsett | | #5

    >"Using natural gas to run a home generator that then runs your heat pump is probably going to be less efficient than using grid electricity to run your heat pump, for example, even if the grid was being supplied by natural gas power plants."

    That's going to be true with HUGE margin. A best-case piston-engine generator at utility scale is going to have only 25% thermal efficiency, which at a COP of 3 would only be 75% efficient net-BTUs. Most home-sized backup generators would only be good for about 15% thermal efficiency, thus sub-50% efficiency on a net-BTU basis at a COP of 3.

    A home sized heat & power cogenerator can hit a net efficiency of about 75%-80% even without leveraging it with a heat pump, but despite great promise a decade ago, there aren't a lot of micro-cogens available in the US market that will work as a backup generator. My biz-partner has been running with a ~1kw Honda cogenerator (net metered) integrated into his heat & hot water system for about a decade, and it runs pretty much 100% of the time when outdoor temps are below 40F or so. Honda sells a similar unit in Japan as a standalone backup generator (complete with a lawnmower-type rope-pull start), but it's not big enough to run a heat pump. They had a hard time keeping up with demand in the first year after the Fukushima disaster, when all nuclear power in the country was turned off. Somewhat bigger micro-grid cogenerators can still make sense in larger buildings (say a NYC skyscraper) with significant thermal loads, with reasonably designed thermal storage systems.

    1. Expert Member
      Zephyr7 | | #7

      I usually allow about 16% efficiency for small piston generator systems, and it’s worse if they run at less than about 80% continuous load. I’ve often thought about running a cogeneration setup, but only at power outage time since I don’t want the maintenance and wear and tear in my genset if I ran it continuously. It’s really the maintenance costs that kill the economic viability of this at a small scale in my experience. The cost savings with cogeneration for short time periods is poor too, and I’ve looked at that. Any liquid cooled generator is relatively easy to adapt to help with space heating using a heat exchanger.

      I did design a larger (~500kw electric output) cogeneration plant for a customer that had their own on-site oil and gas well. The system was designed to use the gas from the well (the oil was sold off), and to run continuously. It was a very interesting project providing both heat AND cooling by using an absorption chiller. I love projects like that, trying to build systems with as close to zero energy waste as possible. We captured waste heat from the engine coolant, the exhaust, and even a little bit from the generator itself (which helped to heat the room the unit was in). Fun project!

      I wonder if there is a possible market for a small scale packaged system. If such a system was purpose built on a production scale, the economics would be much better. The larger systems tend to all be at least semi-custom. It sounds like Honda tried it and had a hard time in the US market.


      1. Expert Member
        Dana Dorsett | | #8

        Honda never really went for it in the US market. A small start-up in Massachusetts cooked up a couple of systems for using Honda's already-developed cogenerator into both hot air and hydronic systems that also included an indirect water heater for extra-buffering of the thermal output. The thermal output of the boilers & furnaces used in those systems was crazy-large for the loads of average houses, but for large leaky antique houses they were reasonable. (My partner's house is an 1840s vintage 5000' house originally built as a summer place for some fat-cat, with lots of very large double-hung single pane windows, etc.)

        They launched the product under then name "Freewatt", which after a couple of years got sold to ECR International, who dropped it a few years later. Maintenance on the cogen is/was provided by the system vendor, contracted out to third parties. The systems were internet-enabled, and report both engine hours, duty cycle, and kwh/temperature data back to automated systems that flag it when something starts to go off.

        In the Massachusetts market those systems were financially viable due to some of the highest electricity rates in the lower 48, combined with fairly generous net-metering policies by the state legislators.

        About a decade ago the German utility Lichtblick launched a program of centrally controlled house or small office building cogens with somewhat large thermal buffering on site using purpose-designed 2 liter VWs. They called it "EcoBlue", operating as a very flexible distributed "virtual powerplant" for balancing wind farm output. I'm not sure how many were ever installed (or still operating), but it seemed like a good idea at the time/for the times. There isn't a lot of recent press on it, but it still comes up in web searches:

  5. Gary | | #6

    Of course efficiency is only part of the discussion, too. The amount and location of pollution is critical to the equation. Centralized fossil fuel use allows use of sophisticated pollution controls that aren't possible, practical, and/or cost-effective at a small scale. And of course emitting that pollution away from population is far better for health.

Log in or create an account to post an answer.


Recent Questions and Replies

  • |
  • |
  • |
  • |