Trying to Predict Gas Prices Over the Life of a New Furnace

When you buy a furnace or heat pump, you're making a 15-20 year bet on fuel costs. That's how long well installed equipment lasts. If you choose a hybrid system — a two-way AC (heat pump) paired with a gas furnace — you get fuel choice. If gas stays cheap, lean on the furnace. If gas spikes, lean on the heat pump. Both go crazy? Install solar and make your own electricity. You don't have to predict the future. You just have to not get locked into a fuel you have zero control over.

I know this because I tried to predict gas prices in the past. I was mostly right about the direction and mostly wrong about the timing. Here's the story, and why it matters if you're buying HVAC in the next few years.

The Predictions
In 2012, I was 34 and running an insulation company in Cleveland. Gas prices had recently spiked — they'd been very expensive through much of 2005-2008, and I was (idiotically) talking to homeowners about payback periods for insulation at those prices. I wasn't an energy analyst. I was a guy who read a lot and could do basic math.

I wrote a blog post called "Why Your Gas Bill May Triple in the Next 5 Years." I cockily called one section "Margins Like a Drug Cartel" — because Asian wholesale natural gas was $20/MMBtu while US prices were $2/MMBtu. A 10:1 ratio. I listed five reasons prices were headed up: electricity generation switching to gas, vehicle fleets converting, well owners capping wells because they were losing money at $2, exports coming online, and growing demand from China and India.

Did I predict $12/mcf retail within five years? I did. What an idiot. My natural gas export terminal timeline: 2015-2016.

In 2014, I followed up with "Oil Is Dead, and It's No Big Deal", where I wrote about Russia threatening Europe's gas supply and predicted we'd have "world priced natural gas in the near future, which could be 2-8 times more than we are paying now." The 2-8x framing left some room. The "near future" part... did not.

I said five years. It took twelve. The export terminals came online in 2016 — which was technically in my predicted range — but prices didn't respond the way I expected.[^1]

Math wins, but it takes longer than it should.

What Actually Happened
The first LNG export cargo left Sabine Pass, Louisiana on February 24, 2016. By 2023, the US was the largest LNG exporter in the world. By 2025, we were shipping 15 billion cubic feet per day overseas — more gas than all 74 million American households on the gas network consume combined.[^2]

Russia did exactly what I worried about in 2014. They invaded Ukraine in 2022, Europe panicked, and suddenly every available LNG cargo was headed to Rotterdam instead of Tokyo. Henry Hub tripled from its 2020 low of $2.03 to an annual average of $6.45 in 2022 — and peaked at $9.68 in August of that year.[^3] That 10:1 ratio I wrote about in 2012 compressed violently and fast.

Prices came back down in 2023 and 2024 — so far down that 2024 set a record low at $2.20. People forgot. They always forget.

What About Iran?

But the structural problem I identified in 2012 hasn't gone away. The Iran conflict is going on as I write this and while US Henry Hub prices haven't spiked too much yet because of US production capability, but 17% of world LNG exports are offline, so what will come of that as we hit winter usage later this year?

Winter future prices are expected to be 50-70% higher than they are right now. Aince the US is able to export more and more to the world, we're likely to be tied more tightly to world prices as we've seen since 2016, that article and a number of other sources say the same thing.  

Bottom Line

Relying on gas, propane, or oil for heating alone is likely to be more expensive during the 15-20 year lifespan of your new HVAC system. 

Let's look at some data:

US LNG Exports & the Link to Your Gas Bill

How America became the world's largest LNG exporter — and what it means for home heating costs

Chart 1: US LNG Exports vs. Global LNG Trade

Billion cubic feet per day (Bcf/d) · 2010–2030 · Shaded region = forecast

Who's consuming all this gas? (2025 US consumption: ~92 Bcf/d record)

Electric power: ~36 Bcf/d (39%) Industrial: ~26 Bcf/d (28%)

Residential + commercial: ~22 Bcf/d (24%) LNG exports: ~15 Bcf/d

Power generation is the biggest consumer — more gas than all homes and businesses combined. And it's growing: data centers for AI are adding 3+ Bcf/d of natural gas demand by 2030 according to industry estimates. PJM (the grid operator for 13 eastern states including OH, PA, WV) has greenlit 50+ new gas-fired power plants to meet data center load. Every new gas plant locks in 30+ years of gas demand.

LNG export terminals already consume more gas than all 74 million US households on the gas network. By 2029, export capacity is on track to roughly double. These terminals operate on 20-25 year contracts — they're not going away.

The forecast: US export capacity is on track to reach ~29 Bcf/d by 2029, with 35 Bcf/d approved. Global LNG trade is projected to hit ~75 Bcf/d by 2030. The US share grows from 27% today to 30%+ by end of decade. Some analysts (IEEFA) see a temporary oversupply in 2027-2028 that could push prices down briefly — but that glut may discourage new investment, setting up a tighter market after 2030.

Chart 2: Henry Hub (US) vs. TTF (Europe) vs. JKM (Asia) Gas Prices

$/MMBtu annual average · 2000–2025 · Henry Hub from 2000; TTF/JKM from 2010 · Note: Annual averages smooth out monthly spikes (Henry Hub peaked at $9.68 in Aug 2022)

The domestic spike era (2000-2009): Henry Hub hit $8-9/MMBtu in 2005 (Hurricane Katrina) and 2008 (commodity supercycle) on purely domestic factors — US gas had zero connection to world markets. The shale revolution crashed prices to $2-4 by 2010-2015. This was the era Nate was selling insulation in Cleveland, talking about payback at $8 gas prices.

Before LNG exports (2010-2015): Zero correlation with world prices. Henry Hub sat at $2-4 while TTF and JKM were $8-16. That 10:1 ratio is what Nate wrote about in 2012 — "margins like a drug cartel." US gas was cheap because we couldn't export it.

After LNG exports (2016-present): The lines started moving together. The 2022 spike is the smoking gun — when Europe's TTF hit $40 on the Ukraine shock, Henry Hub tripled from its 2020 low to $6.45 (peaking at $9.68 monthly). When global prices fell in 2023-2024, Henry Hub fell too. Going from no correlation to some correlation is the key shift. Combine that with rising demand from exports and data centers, and the long-term pressure is upward.

The Bottom Line for Home Heating

If you're buying HVAC equipment, get a hybrid system. A two-way AC (heat pump) paired with your existing furnace gives you fuel choice. If natural gas stays cheap, great — you haven't lost much. If gas spikes because of an LNG shortage in Asia or a cold snap in Europe, you switch to electric heating and ride it out. And you'll probably like your heat pump — they deliver better comfort than furnaces at moderate temperatures, which is most of the heating season.

Natural gas prices face long-term upward pressure. Not because of any single policy or crisis, but because of structural demand growth: LNG exports doubling by 2029, data centers adding billions of cubic feet of daily gas demand, and the electric power sector consuming more gas than all homes combined. EIA forecasts $3.80-$4.00/MMBtu through 2027. The DOE's own study projects a 31% price increase from export growth alone.

Electricity prices at least have a chance of stabilizing or declining — wind and solar are the cheapest new generation in most of the US, and you can install solar to lock in your heating fuel cost for 25 years. You can't do that with natural gas. A heat pump + solar takes you out of the global commodity casino entirely.

The era of $2 gas that made the heat pump vs. furnace debate close is likely ending. Every new LNG terminal and every new data center increases the competition for the same molecules that heat your home.

Data: EIA Natural Gas Monthly, EIA Short-Term Energy Outlook (March 2026), EIA Annual Energy Outlook 2025, IEA Gas 2025,
GIIGNL Annual Reports, IEEFA Global LNG Outlook 2024-2028, FERC LNG Terminal Data (March 2026),
Bloomberg Finance (TTF/JKM), DOE 2024 LNG Export Study, Oxford Institute for Energy Studies. Compiled April 2026.
CSHVAC — natethehousewhisperer.com

Where We Are Now

Eight LNG export terminals are operating in the US. Another 24 billion cubic feet per day of capacity is under construction or approved. If all of it gets built, US export capacity will hit 35 billion cubic feet per day — roughly a third of everything we produce.[^4]

The EIA is now saying the quiet part out loud: "Higher natural gas prices in 2025 and 2026 are the result of strong export growth that persistently outpaces U.S. natural gas production."[^5] Henry Hub is forecast around $3.80-$4.00 for 2026.

And exports aren't the only new mouth to feed. Data centers for AI are adding billions of cubic feet of daily gas demand. PJM — the grid operator for 13 eastern states — has greenlit 50+ new gas-fired power plants to meet data center load. The electric power sector already consumes more natural gas than all US homes and businesses combined — about 36 billion cubic feet per day.[^6]

More LNG terminals. More data centers. Production can grow to meet it — shale wells can be drilled fast — but they also decline fast. Shale wells spike when first drilled and fall off steeply, so you're on a treadmill: you have to keep drilling just to stay flat. The EIA is projecting production growth, but they're also projecting that demand growth outpaces it.[^7]

A Two-Way AC Hedges Your Bet

This is the part that matters if you're buying HVAC equipment: A two-way AC hedges your bet. If gas stays cheap, you haven't lost much. If gas doubles over the next decade — which is what the trendlines suggest — you've dodged a bullet. And if you pair it with solar, you've taken yourself out of the global commodity casino entirely.

Your natural gas bill is no longer set by US supply and demand alone. It's increasingly influenced by cold winters in Europe, economic growth in China, wars in the Middle East, and how many data centers get built in Virginia. Electricity prices aren't immune — gas sets the marginal price in many power markets. But electricity has something gas doesn't: you can make your own. A solar array locks in your energy cost for 25 years. Try doing that with natural gas.

Where I Might Be Wrong (Again)

There's a massive wave of new LNG export capacity coming online worldwide between 2026 and 2028 — the largest in the industry's history. IEEFA projects that global supply capacity could exceed demand scenarios through 2050.[^8] If that capacity hits faster than demand grows, we could see a temporary glut that pushes prices down in 2027-2028. Some analysts project spot prices falling back to $7-8/MMBtu in Europe and Asia by 2029.[^9]
US shale production has also been remarkably responsive to price signals. When prices spiked in 2022, producers drilled more. When prices crashed in 2024, they pulled back. That flexibility has kept domestic prices lower than many people (including me) expected.

So I might be wrong again on timing. Maybe cheap gas has another run in it. But the hybrid argument holds regardless. A two-way AC doesn't cost much more than a one-way AC, and it gives you fuel choice no matter what gas prices do. It's the right bet even if gas stays cheap — and it's a lifesaver if it doesn't.

The Lesson

I was early in 2012. Being early in markets is the same as being wrong, except you get to be smug about it later. Or just feel dumb. Or both. 

But the direction hasn't changed. The US is connecting its domestic gas market to world prices, permanently, through concrete and steel terminals that will operate for 25+ years. Every new terminal ratchets the connection tighter. Every new data center adds demand. The macro trend looks more likely to go up than down.

In 2012 I closed with "these ARE the good old days" for cheap natural gas. Fourteen years later, $2 gas came back one more time in 2024 — a gift from mild weather and overproduction. I don't think we'll see it again.
Math wins. It just takes longer than it should.

Nate Adams is a building scientist and author of Common Sense HVAC, a homeowner's guide to getting a quality HVAC installation. He lives near the New River Gorge in West Virginia, where his gas bill is mercifully low — because he heats with a heat pump.

Sources:
[^1]: US LNG exports grew from 0.5 Bcf/d in 2016 to 15.0 Bcf/d in 2025. EIA, "Ten years after first Sabine Pass cargo," Feb 2026. [^2]: Canary Media, "The US is exporting huge amounts of natural gas," Feb 2026. [^3]: Henry Hub annual averages: $2.03 (2020), $3.89 (2021), $6.45 (2022), $2.57 (2023), $2.20 (2024). EIA Natural Gas Monthly. [^4]: FERC estimates 35 Bcf/d of total US LNG export capacity (14.6 existing + 23.7 under construction + 11.4 approved). American Action Forum, March 2026. [^5]: EIA Short-Term Energy Outlook, Jan 2025. [^6]: US electric power sector consumed 35.8 Bcf/d in 2025 (39% of total). Residential + commercial combined: ~22 Bcf/d. EIA, "Electric sector gas use fell 3% in 2025," March 2026. [^7]: EIA Short-Term Energy Outlook, March 2026. [^8]: IEEFA, "Global LNG Outlook 2024-2028." [^9]: Oxford Institute for Energy Studies, "A New Global Gas Order? Part 1," July 2023.