Hacker News Reader

Comment by dragontamer | Hacker News Reader

Natural Gas heating is surprisingly good in practice, especially in the days where we only had fossil fuels on our electricity grid.

If you go fossil fuel, you have only 40% efficiency for Fossil Fuel -> Electricity, then Electricity -> Traditional heat is only 100% efficient. So 100W of chemical energy turns into 40W of electricity, and then turns into 40W of heating.

---------

Today, a Heat Pump can be like 200% efficient, so you 100W of chemical energy turns into 40W of electricity, then turns into 80W of heating from the heat pump.

Alas, a Gas Furnace is like 90% efficient (10% of the heat escapes in the steam / waste products that needs to be pushed out the chimney, but everything else turns into home heating). So your 100W of chemical energy turns into 90W of heating, and you're done. And that's why our infrastructure in the USA is so much around gas heating, because its better. Especially since we are very rich in domestic natural gas production.

Yeah, we're getting to the point where Heat Pumps + Solar Electricity are coming. But... we're not there yet. On today's grid, Natural Gas heating is likely the most efficient option.

EDIT: Got my units wrong. It should be 100J of energy, not 100W of energy. Watts are power. Though in USA, we don't use Joules for energy, but instead "BTUs". Whatevs.... I think my point is still clear :-)

View on HN

You're at the low end for heat pump efficiency, but more importantly, you're overlooking the fact that the fossil fuel usage of heat pumps benefits from the ability to use electricity generated by non-fossil fuel sources. In the US that's a significant fraction of electricity generation and in many places it's the majority of electricity. The ability to combine the electricity produced by chemical energy with nuclear, hydro, solar, wind, etc, gives heat pumps an significant advantage over directly burning fossil fuel for heat if what you care about is fossil fuel usage.

Renewable fuel sources are already nearly 100% in use. So if you add electricity demand, that new demand usually comes from fossil fuel. That is, you need to consider marginal electricity generation.

At worse, it’s replacing natural gas with fossil fuel. At best, it increases demand so more renewables are built.

Heating and AC are unlikely to be powered by renewables, because they're highly correlated and bursty.

Ex: Your town might be fine with 100MW of power at 60F. But if the temperature drops to 40F, everyone turns on their heaters more, so the entire town uses 150MW or more. So how much renewables do you build? Not 100MW, but 150MW. Which is absurd, no one is going to overbuild capacity like that. And even at 150MW, if the temperature drops to 20F or 10F, even more energy will be used.

In contrast, the Fossil Fuel / Natural Gas peaker methodology is to overbuild on capacity, because capacity is extremely cheap. Maybe the town has a 200MW plant for example. You only use what the people need, so the 200MW plant consumes 100MW at 70F, 150MW at 40F, and everyone has capacity left to grow and expand the town.

----------

Renewables are great for baseload / constant loads that are predictable. But weather isn't always predictable. In this example, 100MW of baseload renewables would be ideal, and the gas-peaker plant is used for everything in addition to that.

You're drastically overstating the case. No, it does not make sense to treat the entirety of heating as non base load (with your anchor of "60F").

On the scale from predictable to not predictable, heating loads are somewhere in the middle. For a given night that ends up having a low of 10F, it would be hard to know that exactly ahead of time, but we have a pretty good idea that it will be below 20F. So that's roughly only (65-10)/(65-20) - 1 = 22% additional heating load that wasn't predictable.

Plus HVAC is highly dispatchable - if there is excess electricity that needs to be used, heating an X-to-water heat pump's storage a few extra degrees is a perfect place to use it.

You're dramatically underestimating the swing of electricity in practice.

https://www.caiso.com/TodaysOutlook/Pages/default.aspx

See September 22, 2022. California required 51,000+ MW of power capacity as the state reached record high temperatures. Just a few days earlier on Sept. 9th, 42,000 MW was all that was needed.

See Yesterday: March 31st 2023. California only required 26,000 MW of power capacity. Even then, the low for March 31st was 17,000 MW at midday.

Given these stats from the last year... How much baseload should California build out? And how much "peaker" production should California have?

--------

The weather swinging +/- more than expected can cause 10,000MW (or 20% shifts) in power in just a couple of days. Seasonal differences are +100%/-50% differences in power capacity.

Its because of this that natural gas peaker plants are used. That's why H2 fuel, as inefficient as it is (but a potentially green source) is still being considered. Having day-to-day, or seasonal, storage of power through the use of chemical energy storage is one of the most promising technologies to tackle this problem.

You've switched from talking about heating (presumably in climate that needs a significant amount), to talking about cooling California.

I'm not versed in exactly how these numbers shake out to be making informed conclusions, but by your previous argument it doesn't seem like you are either. For example cooling load is going to be positively correlated with solar output. More bursty renewables (solar, wind) means less base load and more peaker plants to handle their droops. And just because a town builds one "200MW peaker plant" doesn't rule out adding renewables (and making the peaker burn less gas), even though the renewables only operate at their full capacity a small part of the time. The only way to figure out what to make sense is to tradeoff specific capex/opex/utilization/revenue, rather than talking in wide-sweeping generalizations.

Storage is the general solution to the problem, yes. The easy form of storage that we've become accustomed to is the chemical energy of natural gas. Another is end-user thermal storage, where consumption is accelerated/delayed depending on what the grid needs. Moving away from fossil fuels requires figuring out storage, especially how to utilize the inherent storage/slack we already have, rather than continuing to just take fossil fuels' convenient advantages for granted.

> doesn't rule out adding renewables

I never was against adding renewables. I'm simply pointing out rough estimates on our electricity usage throughout the year, and throughout the day.

Peaker plants, unfortunately, seem like the best solution. Natural gas is the best peaker plant for a variety of reasons. But then natural gas piped to a house for near 100% conversion efficiency is even better (rather than taking rather dramatic hits on conversion inefficiencies). Therefore, Natural Gas in homes is surprisingly useful, and will remain so for the foreseeable future.

We absolutely should be building renewables for the near and moderate terms. We don't have enough renewables for baseload yet.

> You've switched from talking about heating (presumably in climate that needs a significant amount), to talking about cooling California.

California has public electricity usage patterns. For better or worse, its the best data that we've got. I assume its somewhat representative of the energy requirements across the country.

Other parts of the country its way more complex because there's natural gas flowing everywhere. So its harder to convert Natural Gas heating vs electricity prices. Nonetheless, the concept of "This differs based on the weather" and "The weather changes dramatically every week" is universally true.

I wasn't saying you yourself were against renewables. Rather your argument seemed to imply that building one big peaker plant would preclude the demand for renewables since the peaker could supply the entire load all the time.

I don't think we'll ever have "enough" solar+wind to supply base load, since night exists and winds calm down. Talking about "renewables" as a group doesn't really make sense. For example hydro is highly consistent and highly dispatchable, but waterfalls are rare.

Spitballing, I would think cooling (especially with California's coast where it's rarely needed) would have much higher variability than heating, since people seem to tolerate a wider range of higher temperatures compared to colder temperatures. At least from personal experience once I enable the heat for the year it mostly stays that way, turning on when it gets even a degree too cold. Whereas air conditioning is a question each day of "can I get by with the windows open", and only when the answer turns out to be an unequivocal "no" do I turn on the AC (which then runs continuously to make up for lost time).

Natural gas is a fossil fuel

We need to go to zero and gas furnaces cannot. (Horribly inefficient synthetic fuels excluded.)

In that context current electricity generation does not matter. It‘s an infrastructure change that allows for future zero emissions. Just like EVs.

We're at like 12% renewable. The journey from 12% renewable to 30% renewable will be different than the journey from 90% to 100% renewable.

The tools in our disposal to reach 20% or 30% are available today, if we can understand them. And also, as long as we don't let perfection be the enemy of good.

We will probably never tackle the 90% to 100% problem in our lifetime. It is best to focus upon the problems that are in front of us, and that we can in fact solve or fix within our lives.

We need to get to zero. Burning gas doesn’t get us to zero. That’s the issue.

Given the lifetime of gas furnaces and what we know about climate change it is madness to even consider using them in brand new construction now.

Gas was and is the right tech for transitioning. But you seem to be stuck in the 90s or early 2000s when that kind of policy would have made sense. It’s too late and getting later.

>Today, a Heat Pump can be like 200% efficient, so you 100W of chemical energy turns into 40W of electricity, then turns into 80W of heating from the heat pump.

This isn't true today. The absolute minimum COP is 3.1 in the US.

https://www.pickhvac.com/heat-pump/basics/cop/

This appears to be the minimum for geothermal. The vast majority of heat pump installs would be air source, and while these can often get 3+ it’s highly dependent on climate and local grid efficiency.

It also depends on local prices. In Madison, WI our implied (not actual) electric grid efficiency based on prices is under 20%.

The energy.gov page has more details, it's 3.3 for air source at 47 deg F and a reasonable size:

https://www.energy.gov/eere/femp/incorporate-minimum-efficie...

Not for air source.

Source: your link.

Air source heat pumps are a lot easier, so they are the majority of installs.

I expect the national code doesn't give a minimum COP for air source because the performance changes with air temperature. The bulk of my heating cost occurs when it is 10-20 ⁰F outside, where someone located in a warmer spot might have most of their heating at 35-45 or whatever.

I definitely read it too quickly, but clicking through to the energy.gov page does clarify that the COP is measured at 47 deg F (which is warm)

You will not get the COP of 3 during most of the winter in the interior US with an air source heat pump. Their efficiency go down significantly once the temperature is below freezing, due to required thawing cycles, and because the bigger delta between inside and outside temperatures reduces pumping efficiency in general (efficiency is best when outside is not much colder than inside).

Air source heat pumps are really good tech overall, especially because they also double up as AC in summer. However, in winter, in a head to head comparison, they only handily win in areas with mild climate, like eg. all west, or much of southern US. In the northwest or midwest, they are unlikely to beat efficient gas furnaces.

Here's a heat pump that has a minimum COP of 2.93 at 5 F, and 9.42 (!!!!) at 47 F: https://ashp.neep.org/#!/product/56885/7/25000///0

Your point stands, but your 1,500 sqft home needs about 68,000 btus.

That heatpump would be for about a 300-400 sqft room.

Seems like a bigger one would only be more efficient.

Surprisingly, this doesn’t seem to be true for most of the split system ductless heat pumps (aka “reverse cycle air conditioners), I’ve seen.

The smallest ones are almost the most efficient ones in the range.

I’m not sure why this is, and whether it generalises to house-size ducted systems, but it’s pretty striking nonetheless because it’s the exact opposite of most heat related systems I can think of.

Might it be because the surrounding air only has so much heat in it?

Thus, smaller units, more spread out, have access to more heat?

I don’t know why - But your single air exchange heat pump driving a single head unit in the 9,000-12,000 range seems to be incredibly efficient. Once you go to a larger unit, and more heads, the efficiency drops.

It’s fascinating to even see the huge increase in switchable units - running your 4/5 ton unit at 4 tons is much more efficient. (Usually)

If you look at the numbers that's clearly not the minimum COP. I'm not sure what it is, I think it's the COP when operating at minimum power.

Look in the right most column for COP at max power

Cold weather heat pump technology is consistently getting better, but I also think people underestimate how many heating days are above freezing even in cold areas and how much of the US and other places is mild enough in the winter for heat pumps to be fairly efficient nearly every day during the winter (especially if you look at population rather than land area).

If you look at a place like Minneapolis, which is normally considered a pretty cold US metro area, you do have a few months where the average temp is below zero. But you also have several more months where the temperature is cold enough to turn on the heat but well above freezing, where a heat pump is dramatically more efficient. When picking a heating system, it makes sense to consider comparative efficiency over the entire year, not just performance during the worst conditions in the middle of winter. Even in a city like Minneapolis, it's not obvious that heat pumps are less competitive over the entire year, even if their performance in January probably isn't fantastic.

I also think the practical geographic area extends further than people think. New York City definitely isn't the south, but per some random Googling, it doesn't appear to have any months where the average temperature is below freezing. Even Boston only has 2 months that barely drop below freezing. Most of the US northeast (definitely the mid-atlantic) by population arguably also falls into the range where heat pumps make sense.

But then you need two systems - heatpump for most of the year, and natural gas for the cold days when you would die without it.

At least in Canada I’d need to know that my system can handle a week of -30 or -40C in my city. Some places are even colder, for longer at a time.

No, electricity is versatile, you can just make electricity into heat on that day when it's too cold for your heat pump. Resistive heating is not exciting > 100% efficiency technology, but it works and if you live somewhere that realistically gets too cold to run the heat pump that's what you add, not a whole gas burning system.

I don’t know why I didn’t think of that.

I guess the only remaining concern is that winter storms cause both cold weather and trees falling onto powerlines. The latter is rare though

Right, but (a) most modern gas fired heating actually requires mains electricity to function properly, so if you are blacked out your heating probably is disabled and... (b) even if the heating does work, a modern home is a pretty miserable place with no electricity, unless you know power will be back soon (hours or sooner) you should probably set out for a communal safe respite centre where everybody can work together to stay warm, feed and care for each other until the emergency is resolved.

There are people who genuinely live in the middle of nowhere, on an otherwise uninhabited island as caretaker, deep in the forest manning a fire watch tower, Antarctic research base, that sort o thing, who do need have more options, because if they run out of options they're dead. If you're in that situation you still don't necessarily want gas heating though, a fossil fuel powered generator might make a lot of sense, maybe even more than one, because again electricity is so versatile.

A snow storm plus power outage makes me want to bed down, not flee. The idea of leaving your house for a community center during a power outage is odd to me on several axes (pipes freezing, digging out your car, driving on half-plowed roads), especially in the hyper-individualist US (do these centers even exist?)

It's much simpler to obtain and fuel a small generator to power meta loads like thermostats/circulators/blowers than a large generator that supplies the thermal load. Fully committing to a heat pump as your only heating would likely involve a large gas generator fed by municipal gas or at least a large buried propane tank (for capacity and temperature).

The straightforward backup heat answer is wood. It's easy to obtain and store, and the primary drawback of it requiring labor and attention isn't a big deal during an outage. Although if you don't do it regularly and make sure your stove is in good order, lighting a fire during an emergency situation might not be the best idea.

So true, I'd much rather batten down the hatches, put my frozen food outside, and make sure none of the pipes burst.

I heard about a guy who had copper piping with water flow going around his chimney (on the inside) and then through the house.

That'd be a cool system if you could switch out the heat source, electric and an indoor-safe wood stove. Just get a CO monitor!

Actually, I just got a wood boiler for this past winter. This is basically a stove that lives outside, surrounded by a water jacket (~200 gallons) and insulation. It modulates the fire to keep the water from boiling off regardless of how much heat you're drawing off. You then run pipes to your house and tie it into your hydronic heating system. I feed it wood twice a day, and it keeps my house as comfortable as burning oil did (actually even more comfortable since I can heat the unfinished basement when I want, and not feel bad). And since it's outside, I don't deal with any of the mess or work of bringing heating-scale quantities of wood inside.

They do require power to run (automated dampers, blower, circulators), so I have a small backup generator as per my last comment. And they've gotten a bit expensive (especially now that the US federal tax credit has gone away), but my estimated payback period is around 6 years.

You might wonder why I'm posting in a thread about heat pumps when I just got a different major heating source, but I see heat pumps and wood as two very complementary heat sources. Down the line I'm planning to add a water-to-water or air-to-water heat pump for the shoulder seasons (and summer DHW), so that I don't have to tend a fire (or burn oil) when the heating load is low. One four or five ton unit should take care of my heat loss most of the time, and I won't have to oversize the system to handle the dead of winter cold snaps.

The only problem with that is if your electricity generation capability is correlated with the outside temperature. You don't want everyone turning on their resistive heat right when you have the minimum generation capacity.

I can’t reply to telaramex directly but backup electric resistance is exactly how my heat pump hot water works.

Geothermal is kind of amazing. We have a geothermal system that can get to 5.1 COP. We used 650 kWh for heat in December, which translates to 22 therms. The average household in my area uses roughly 110 therms per month for winter space heating.

Yup, I have a ground source heat pump, sometimes called geothermal - it is amazing! My neighbor has a slightly larger house, but close. Their bill one month was almost $500 and mine was $110.

It didn't cost that much more to install and there isn't any equipment outside in the weather.

My understand was geothermal installation is extremely costly due to the deep bores you need to dig. Or do you have enough land to go out instead of down?

Can you or the GP share more about the economics of your installations?

I've build 3 houses. For the first I had to get a bid for 3 bore holes, that was too much for me. I opted to insulate like crazy and use natural gas.

I didn't bother with a bid the second time because the situation was the same. Air source was really coming along at this point (at least in the US), I should have pushed harder to get it.

My current home has enough land to go horizonal. There are 4 trenches, 6 feet deep, 3 feet wide and 300 feet long. The pipe loops down and back on each trench. Excavation was about $1000 (full day for a man and a machine) and another $1000 to bury them and a day of labor to lay the plastic pipe and weld it together.

Equipment was about $8,000 more than my other options (no natural gas here), but after tax breaks it was about $3000 more. I think it was worth it, even if there weren't tax breaks. I really had to push hard to get this. I had to use a general contractor on this house and he swore it wouldn't work. I made the calls and found someone to do it.

The system runs often when the highs are -20f, but I've never flipped on the backup electric resistance heat.

During the summer the A/C is so cheap it is practically free. It doesn't take much compared to heating. Cooling might have to get the house 30 degrees less than the outside air, but heating has to raise the house nearly 100 degrees f warmer than the outside air.

Ground source is always my first recommendation, but air source should be more popular than it is.

A funny note about the builder who thought ground source wouldn't work well, he also tried to talk me out of a tankless hot water heater because the well water would be too cold in the winter! I guess he didn't understand that the ground down far enough is a pretty steady temperature.

> During the summer the A/C is so cheap it is practically free. It doesn't take much compared to heating. Cooling might have to get the house 30 degrees less than the outside air, but heating has to raise the house nearly 100 degrees f warmer than the outside air.

You can't rely on just comparing temperature differentials. During summer, the house receives a lot of extra heat in the form of sunlight.

It is true that air temps alone do not tell the whole picture, but my coldest winter days also tend to be clear skies and also have lots of direct sunlight. My summer days are also more overcast than one might expect.

It isn't my profession, but I've spent a fair amount of time tweaking building envelope and conditions in HVAC calculations to make decisions in how I build.

A delta of 30 vs 100 is not the only factor, but I'd bet in my case it is the largest factor. I still agree with your point of bringin up other factors. Even changing the direction a building faces can change the load on it.

The US Code vs the Carnot's theorem, who would win? The answer might surprise you!

Got all your numbers wrong too, not just your units. Modern combined cycle power plants mean that 100 J of chemical energy turn into more than 40 J of electricity (depends on your energy mix, but could be as high as 60 J). Then you also get a third of your energy from nuclear in some places, or hydro, or solar and wind (that is happening now - not just in the future). Heat pumps can be 300% or 400% efficient for most of the year (it's those few coldest nights of the year where they really struggle).

So even if you don't have any nuclear or renewable energy in your mix, just natural gas, you get 60*3 or 4 = 180 to 240 J of heat energy from your 100 J of chemical energy.

For most people in America, heat pumps are currently the most efficient option.

> If you go fossil fuel, you have only 40% efficiency for Fossil Fuel -> Electricity,

This is a big underestimate for natural gas now that we mostly use combined-cycle gas plants (basically a jet turbine bolted to the ground, with its exhaust heat used to power a traditional steam plant). Combined-cycle efficiency is over 60%, which changes your math significantly.

more importantly, decoupling the electricity generation with the heating, means that improvements made to electricity generation instantly flows into each home, vs using gas for home heating means _every home_ would need to purchase new gas equipment if improvements become available, to get the improvements - a major capital cost for the individual which may not be able to do it.

Ant then you have combined cycle + district heating plants, and you use even more of the energy.

FWIW, modern heat pumps can exceed 400% efficiency, so the math is starting to work out in their favor even for fossil fuels burned at a power plant.

> Alas, a Gas Furnace is like 90% efficient (10% of the heat escapes in the steam / waste products that needs to be pushed out the chimney, but everything else turns into home heating).

And there exists even more efficient options today. Like high efficiency furnaces that use the heat from the flue in a second heat exchanger to bring it closer to 99%

But having a furnace means you also have central air. So the question isn’t furnace or heat pump, it’s furnace and one directional heat pump or just a heat pump. Yes, on cold days the furnace will be cheaper, but on most days the heat pump is cheaper and requires less maintenance since you’re turning two appliances into one.

Many places in heating-dominated climates do not have central air.

I’ve lived 35 years in New England and only one of those places (a modern high-rise) had central air.

Are we talking window unit plus radiator or just no cooling? For window + radiator heat pumps make less sense, but window unit heat pumps work well and are worth considering as a radiator replacement when the window unit is replaced.

Many have no cooling at all. Others have window shaker ACs only in bedrooms (which aren’t practical to use window heat pumps as a replacement for radiators since the heating balance would be terrible with the rest of the house and the air leakage around a window unit is really bad when the inside/outside delta-T is 60+°F in winter [versus ~15°F in summer])

Window units in practice are "no cooling until there's a heat wave half way through July that prompts you to actually install them".

Uhm. The entire midwest is almost as cold as New England (or colder) and yet has widely installed A/C to deal with hot humid summers.

Modern (i.e. non-adobe, non-stone) construction in the higher-altitude mountain west has similar issues (without the humidity).

New England is really the exception to the rule, not the rule (it's a lovely, lovely exception though).

It was a bit weird when I lived in SoCal. It was the only housing I've lived in without AC. It only got uncomfortable for about a week out of the year, most of the time the ocean winds cooled the air off enough. I can't imagine going without AC living in the midwest and the winters can be absolutely brutal on top of it.

A modern natural gas power plant has over 60% efficiency of electricity generation. 40% is for rather old ones or those used for peak, not base load. But even those are improving.

Moreover, using natural gas at home for heating is not 100% efficient. Some heat is lost with exhaust plus not everything is burned. A good furnace with proper heat exchanger is about 90%, but cheaper ones can be below 80%.

So even if your electrical heat pump is just 200% efficient and the electricity comes from a gas power plant, it is already a win compared with a gas furnace at home.

Both with electricity and gas there are transmission losses, but they are roughly the same and with electricity we know how to make them smaller long-term.

You're focusing on "today" when this is a twenty or thirty year machine. Twenty or thirty years from now a gas furnace will still be a 100% emitter. A heat pump has the ability (without upgrades) to move an increasing quantity of the energy to renewables.

Most traditional natural gas heat is only 80% efficient since you have wasted exhaust heat. If you had a air pump you can achieve up to 92%-95% efficient, but with the added cost and complexity of having an electric fan.

You lose a lot of heat due to improper insulation or in the case of northern california, no insulation whatsoever.

Heat pump's coefficient of performance is more like 3 to 5.

Reply on news.ycombinator.com