it doesn't generate much in terms of particulate emissions (eg it's not dirty like diesel) but it's quite bad in terms of greenhouse effects. the natural gas pipeline network leaks a relatively large amount of the gas that transits it (it's on the order of several percent per year) and because it's such a distributed problem it's not economical for companies to fix it purely on a cost-savings basis. And natural gas (methane) is an extremely potent greenhouse gas - it is #2, behind only carbon dioxide in overall impact on the atmosphere, and something around 100x as potent as CO2 over a 100 year lifespan.

https://www.cnbc.com/2018/07/06/the-us-natural-gas-industry-...

new studies suggest the leak rate is around 2.3%, which is about twice as high as previously predicted - and at 3% you have offset all the benefits compared to coal generation. Again, obviously that's still better in terms of particulate emissions/acid rain, but you wouldn't be doing any less global warming. So it is still significantly bad in terms of greenhouse effect, not nearly the panacea the fossil-fuel industry would like to present it as.

(and conversely - the fact that it's such a distributed problem makes it difficult/expensive to clean up. you can capture CO2 and particulates from a point source - it's expensive and it sucks but you can do it. But you can't stop 2 million miles of pipes from leaking a little bit at every fitting.)

the only saving grace is that methane does oxidize relatively quickly, on the order of tens of years, whereas CO2 is pretty much there forever. but if we're still pumping out more natural gas into the atmosphere 10 years later, that's not really the fix people think it is. Again, it's 100x more potent over a 100 year lifespan than CO2, the "long term" picture isn't all that rosy either.

slight addendum to the above - I should have said "not dirty like fossil-diesel". I'd like to make the point that synthetic diesels like biodiesel are extremely clean in terms of particulates and is in fact one of the more interesting "bio-fuels" (although of course there are always problems with any approach that puts food in direct competition with fuel for feedstocks).

How much of that gas was going to leak anyway? The fields they pipe it from probably don't hold it all in. Any stats on this in Google are buried under reports on the extraction industry, so I can't find anything concrete.

"Anyway" from what reference point? The gas stayed in the ground for millions of years before we poked a bunch of wells into the reservoir. If the wells are shut in correctly, they shouldn't leak appreciably. As the comment above notes, the more pipes you have, the more gas is going to leak.

Unfortunately, in many places you can't produce oil without producing gas, and with so much gas right now, there's not a lot of financial incentive to capture it all. If we reduce the need for oil, we'll reduce gas leak emissions as well.

That's the question. How much makes it to the surface? Oil used to pop up everywhere before humans sucked up all the easy to reach surface sources. There's still a little that comes up from the ocean floor without any human help.

https://response.restoration.noaa.gov/oil-and-chemical-spill...

Why would it be different for other things created by the same processes? And do the pipes let less leak out than if we just left it in the ground?

This source suggests that geological methane emissions are about 40% of total emissions: https://phys.org/news/2021-03-natural-geological-methane-emi...

Oil and gas seeps notwithstanding, the vast majority of reservoirs are stable and likely to remain so over human timescales unless we stick wells into them and start piping fluids around on the surface.