Talk:Peak gas

US peak

How do US production figures look if we ignore Alaska and the Gulf of Mexico? Looking at the Eia numbers for 2003, if you just look at the lower 48 and remove the Gulf of Mexico and coal gas (which it seems Hubbert did not include in his calculations), then the production for that year comes to 12 tcf/y. That seems pretty close to Hubbert's numbers, and it keeps the geographical area and technology Hubbert was considering. Adding coal gas brings the total to 13.8 tcf/y, so really the most significant factor is the addition of a geographical region. That seems to be a pretty unfair change to include in discussion, considering the fact that Hubbert's models are based on set geographic regions. NJGW (talk) 14:42, 10 September 2008 (UTC)

What you are proposing distorts Hubbert's own predictions. 1) In 1956, Hubbert did not exclude Alaska. 2) Hubbert did not exclude Gulf of Mexico in either 1956 or 1971. 3) Hubbert did not exclude coal bed natural gas. I don't recall that Hubbert ever hedged his predictions to exclude new gas sources or new technological developments; if so, will you please cite them. Again, if a method of prediction proves to have no predictive power, then what good is it? What you are saying is: "Hubbert could have predicted the future if only he had known the future." Well, so can you, and so can I. Plazak (talk) 15:24, 10 September 2008 (UTC)

What is your source for Hubbert including Alaska and the Gulf in 1956/71? I don't recall saying Hubbert excluded new technologies, but above I did give conventional and non-conventional numbers just for comparison. NJGW (talk) 15:37, 10 September 2008 (UTC)

The Alaska North Slope gas reserves (which are the largest in the U.S.) are stranded gas and will remain stranded until someone builds a pipeline to the lower 48, so they haven't entered into the peak gas scenario yet. The recent increases are non-conventional gas going on-stream due to better technology and higher prices. In this sense it is similar to the Canadian oil production curve, which peaked in the early 1970s as conventional oil ran out, declined, and then went back up and is headed toward a much higher non-conventional peak somewhere around the middle of this century as oil sands production goes on-stream. I won't go so far as to predict when U.S. non-conventional gas production will peak.RockyMtnGuy (talk) 16:14, 10 September 2008 (UTC)

"United States" oil and gas production includes Alaska and offshore in the common meaning of the term. If you have information that Hubbert excluded Alaska or offshore in 1956, or excluded offshore in 1971, then it is up to you to cite the source; you have been a wiki editor long enough to know that. If you lack a source, it's just speculation on your part. Plazak (talk) 18:10, 10 September 2008 (UTC)

Let's review the discussion:

Me: "How do US production figures look if we ignore Alaska and the Gulf of Mexico?" (never made a claim on Hubbert's work)

You: "In 1956, Hubbert did not exclude Alaska. 2) Hubbert did not exclude Gulf of Mexico in either 1956 or 1971."

Me: "What is your source for Hubbert including Alaska and the Gulf in 1956/71?"

You: "If you lack a source, it's just speculation on your part."

You have made a claim. Please source it. NJGW (talk) 03:00, 11 September 2008 (UTC)

My sources are Hubbert (1956) and Hubbert (1971). Please read them. You will see that Hubbert did not exclude Alaska or Gulf in his 1956 article, and that Hubbert did not exclude the Gulf in his 1971 article. Why are you so intent on jiggering the data after the fact to minimise the inaccuracies of Hubbert's predictions? Let the facts speak for themselve. While your are at it, why don't you exclude Oklahoma? That might make Hubbert look better. Plazak (talk) 13:26, 11 September 2008 (UTC)

Those are long and unsearchable PDFs, please give page numbers. NJGW (talk) 08:11, 12 September 2008 (UTC)

You are asking me to cite a negative, which, as you well know, is b.s. Nowhere in his 1956 article does Hubbert exclude Gulf or Alaska. Nowhere in his 1971 article does Hubbert exclude the Gulf. Again, please read the articles; it would take you less time to read them than to continue this silliness. Plazak (talk) 12:17, 12 September 2008 (UTC)

No, I am asking you to cite a positive: which areas did Hubbert include, as your claim is that Hubbert specifically includes these areas. You must have a basis for this claim, so I would like to know what the basis for that claim is. NJGW (talk) 16:00, 12 September 2008 (UTC)

Hubbert wrote US in 1956 and Lower 48 in 1971. I took his written predictions at face value, and compared them to US (for 1956) and lower 48 (for 1971) production. Unless you can show verifiable information that he excluded Alaska, Gulf, or coal formations, your assertions are your own speculation, and do not belong in the article. Plazak (talk) 18:48, 12 September 2008 (UTC)

Why do you believe that Hubbert included Alaska in 1956 but not 1971? NJGW (talk) 19:54, 12 September 2008 (UTC)

Did Pratt include Alaska and the Gulf of Mexico in his EUR? NJGW (talk) 20:01, 12 September 2008 (UTC)

You guys are somewhat missing the point that Alaska gas is not being produced to market at this point in time. It is what is known as stranded gas because there is no pipeline to the lower 48 states. The largest gas plant in the U.S. is on the Alaska North Slope, but all it does is strip the gas off the oil production stream and reinject it back into the oil field. So, in quoting the EIA gas production numbers, are they referring to 1) all gas produced (including that reinjected), or 2) gas produced and marketed? Most likely, it's the latter, in which case most Alaska gas would not be counted. When Alaska gas does come on stream, there will be a big spike in U.S. production. However, the recent increases in U.S. gas production have been what is referred to as unconventional gas, much of which comes from shale gas formations, notably in Texas.RockyMtnGuy (talk) 03:12, 1 October 2008 (UTC)

I think the chart currently in the article (NG withdrawals) is produced plus reinjected, and this graph is produced and marketed. Thoughts? NJGW (talk) 04:32, 1 October 2008 (UTC)

That's correct. The chart in the article is "gross withdrawals" which includes reinjected gas, probably including Alaska production, while the one you pointed to excludes reinjected gas, which would exclude most Alaska gas. Interestingly, you get a different picture depending on which data you use. If you look at the latest gross withdrawals data, it is rising to a new peak, having exceeded the 1971 peak in 2007, and monthly data shows that 2008 is going to be even bigger. Meanwhile, marketed production (which excludes Alaska gas) is lower than the 1970s peak. So, a lot more gas is being produced but it's not going anywhere. Interesting, no? I expect that if they build the Alaska gas pipeline, all that reinjected gas will come down the pipe to market in short order. Until then, they could include the same gas in "gross withdrawals" over and over again as they reproduce and reinject it. Between stranded gas and gas reinjection, not to mention unconventional gas, the Peak gas scenario could be a lot more difficult to analyze than the Peak oil one.RockyMtnGuy (talk) 15:26, 1 October 2008 (UTC)

 Done

Hubbert's theory

Let's get back to basics. There are two main aspects to Hubbert's theory. The first is that oil and gas production versus time can be described by the mathematical curves presented by Hubbert. The second aspect, and most controversial, was Hubbert's bold thesis that future oil and gas production could be predicted based on data available to him at the time. Hubbert's assertion of predictive value is subject to validation in a straightforward manner: examine predictions made on the basis of Hubbert theory, and see how accurate they turned out to be. The most accessible predictions, and those with the longest track records, are those by Hubbert himself concerning US oil and gas production - and who better apply his theories than the master himself? The comparison of Hubbert vs. US oil production was already described in the wiki article, so I presented - without editorializing or excuses - factual comparisons of his 1956 and 1971 predictions of US natural gas with actual production. Based on the record of past predictions - although in the article I left it to the reader to draw his own conclusions - it may be that Hubbert curves for gas have descriptive value, but little or no predictive value. Now it may be, as you believe, that later predictions will be more accurate. Rather than dedicating your time to making excuses for failures of past predictions, the best way to prove or disprove the predictive value of Hubbert theory would be to present later Hubbert-theory predictions, perhaps for other regions or countries, that successfully predicted future peaks and declines that were subsequently validated by actual results. After all, it has been 37 years since Hubbert's 1971 US gas prediction; there must have been many Hubbert theory predictions in the intervening years. If you know of some that can be compared to later results, it would be illuminating if you were to put them in the wikipedia article. Plazak (talk) 17:32, 11 September 2008 (UTC)

You seem to believe that the accuracy of Hubbert's predictions in 1956 and 1971 are of vast importance to this arcticle. Why is that? This is an article about the concept of Peak gas, its foundations (in that gas is a finite resource and in Hubbert peak theory), and its implications. Current predictions would be a good addition, but harping over the predictions of 52 years ago seems odd. NJGW (talk) 19:54, 12 September 2008 (UTC)

My question is whether peak gas has real predictive value (as Hubbert asserted), or is it something that can be accurately applied only after the fact. Hubbert's own gas predictions have not fared well. As I say above, if you have any subsequent predictions to add, it would be instructive to see how well they have held up. Plazak (talk) 18:16, 15 September 2008 (UTC)

You're making a basic mistake that is influencing your view of the article: peak gas isn't a prediction, it's an event. The predictions are just a part of the story; this article is supposed to be about the event, not the predictions. You also seem to think there is a basic flaw in Hubbert's methodology. What is that flaw and what source tells you this is so? NJGW (talk) 22:24, 15 September 2008 (UTC)

Hubbert himself (1956 and 1971) considered that he could use his peak theory to predict future oil and gas rates. The success of his 1956 oil prediction is justifiably held up as an example of this (in the wiki article as well as in other discussions of resource peaks). I think that it's a fair and quite interesting aspect of his theory to examine if Hubbert peak theory can provide useful predictions for other commodities; a factual account of Hubbert's gas predictions is a logical place to start. I only ask that you not allow your own point of view to prevent the inclusion of documented facts because you fear that the facts would, in your words, "confuse a reader to think that because Hubbert's 1956 estimate was off the theory is bunk." Facts are facts, and not to be dismissed. You are, of course, welcome to add additional documented facts; that is how really good and balanced wiki articles are created. 52 years after Hubbert's article, there must be other, perhaps more successful, predictions of peak gas; If I knew about them, I would include them; please add any that you know about. As for my own opinion, I am agnostic on the subject of the predictive value of Hubbert; I just don't want to take this, or any other theory, on faith. Plazak (talk) 23:54, 15 September 2008 (UTC)

Why are you still insisting that this article is about Hubbert peak theory? Hubbert is not in the title, only the history of the concept. This is an article about the production curve of a particular commodity. Please see peak oil and peak uranium if you are still confused about how an article like this is structured. NJGW (talk) 00:14, 16 September 2008 (UTC)

NJGW, I believe that Plazak does have very useful point: Hubbert's theory *does* have much more descriptive value than predictive value. In part, Hubbert's theory cannot predict the effects of non-scientific influences such as politics, conservation, deliberate slowing of production such as monopoly behavior, etc. If Hubbert's theory correctly predicted several peaks, it's purely out of coincidence. In part, the article is about Hubbert's theory applied to peak gas. However, it's not just Hubbert's theory, but the refinement of others.

Plazak, I hope to find some of the articles for peak gas, but they are very few. Since you are a geologist, perhaps you can research a bit and give us links to articles you run across.

About the article itself, we just got started here while being distracted by several hostile editors on other peak resource projects: Peak oil and Peak uranium. This article contains an outline of what needs to be added and how the article will be built. I hope that the outline can be followed. Thanks for helping. Kgrr (talk) 07:20, 17 September 2008 (UTC)

 Done

Merge proposal

I propose a merger between Peak gas and North American natural gas crisis. I question whether the latter needs it's own article, and it fits nicely into either this article or Gas depletion. NJGW (talk) 23:20, 22 September 2008 (UTC)

Yes - move into Peak Gas: (It's a rate problem much more than it's a problem with the reserves) "United States natural gas production history shows that new wells are being depleted more quickly all the time; production from the newest wells have been declining about 28% per year. While this is partially due the fact that the large plays of NG are all aging and are in terminal decline. Newer natural gas fields tend to be smaller and are produced (and depleted) quickly in the effort to maintain overall production levels." "In spite of increasing the number of natural gas wells by about a third since 1999, annual production peaked near 50 billion cubic feet per day and now continues to fall, so a growing percentage of natural gas supplies are being imported from OPEC and other countries." Natural Gas Crisis Kgrr (talk) 00:28, 24 September 2008 (UTC)

 Done

Shale gas !!??

To read this article one would not think that dramatic discoveries in shale gas have been made in the U.S. over the past 5-10 years. These new plays use newer horizontal drilling techniques that have only recently been developed. There is now a land rush for leasing rights to use this technology in certain parts of the U.S., particularly NY and PN.

To leave this information off seems a glaring omission. It deserves at least a paragraph of proper discussion. Some now think that we are entering a natural gas glut, in which these new plays will maintain modest prices for the foreseeable future. 12.171.32.226 (talk) 23:15, 4 December 2008 (UTC)pjc

Yes shale gas has contributed to prevent a sharp decline in the US gas. Please feel free to add it under the United States section. There are large quantities of shale gas, but it's not being produced as fast as primary natural gas was. Kgrr (talk) 19:33, 23 December 2008 (UTC)

There appears to be a large amount of shale gas in the US, which has been reversing the historical decline in US production. Industry insiders tell me that the US has about 50 years of shale gas in the ground. Interestingly, they also say that Canada probably has several times as much shale gas as the US. If that's true, then North America has several hundred years of natural gas supply remaining, albeit at higher prices than current ones. Unfortunately, none of the sources are quotable, so it would be tough to put in the article.RockyMtnGuy (talk) 05:50, 24 December 2008 (UTC)

Thanks for the info, RockyMtnGuy I would suspect that shale gas could probably increase production in other countries besides the U.S. and Canada, perhaps to dramatic levels. To my mind, the information on shale gas alone is enough to allay any concerns re: Peak Oil. It is a reliable energy source that can be easily used for most industrial purposes. There are vast quantities that can be extracted for a price that is, while not exactly cheap, is not particularly exorbitant either. To my mind, shale gas is the equivalent to finding large quantities of domestically available $70/bl oil. Not a windfall, but a sufficient hedge against disaster. 76.105.219.46 (talk) 03:11, 17 March 2009 (UTC)pjc

There are a few problems with that:

1. NG is not as easy to use as a transportation fuel. While you can run a vehicle on NG, it requires replacing the entire fuel system and putting a CNG tank in the trunk. If you want to use CNG, you need access to a compressor.
2. NG is not as transportable as oil - if you're not on the pipeline network it's difficult to get it.
3. At current prices, it's not economic to drill for NG, so companies are suspending their drilling programs, leading to a future decline in production.

So, it's not equivalent to finding large quantities of domestic oil - which in the case of the US is nearly all gone.RockyMtnGuy (talk) 21:57, 17 March 2009 (UTC)

My point is that if the price of oil were to remain above $70/bl for an extended period of time, it would likely spur a shift to use NG instead of oil for some applications, which would in turn drive up the price of NG (and check the rise in the price of oil), which would in turn tap the shale gas in the U.S., Canada, other places. It seems hard to imagine the world reverting to a "pre-technology" culture (as the Peak Oiler maintain) while there are still so many energy reserves that remain untapped. 12.171.32.226 (talk) 18:44, 19 March 2009 (UTC)pjc

US and North America

The United States does not take up all of North America. When adding material under the United States as a country, references for peak gas in North America are not exactly relevant since peak gas has already happened in the US in 1973. The gas in North America that has yet to peak is in Mexico and Canada. I will find some time to clean-up this confusion. Kgrr (talk) 19:33, 23 December 2008 (UTC)

Thanks for the geography lesson. However, peak gas in North America is relevant to the US in the sense that the US gas market depends to some degree on imports pipelined in from Canada. The ref that you are looking for is already in the article: US Energy Information Administration: World dry natural gas production Retrieved 7 December 2008. Plazak (talk) 19:45, 23 December 2008 (UTC)

The US accounts for about 2/3 of North American production. US production appears to be headed to a new peak because of increased production of shale gas, but I think that is somewhat dependent on prices, which are falling. Canadian production peaked in 2002 and appears to be undergoing an irreversible decline. Mexico is not a significant exporter of gas, and in fact seems to have trouble meeting its own internal demand.RockyMtnGuy (talk) 05:33, 24 December 2008 (UTC)

The oil production data for the US and Canada does not support a new peak. US production in 1973 was 24.1 Tcf, in 2001 it was 19.6 Tcf and has dropped since to 18.5 Tcf. There has not been an sharp increase in total US production as you claim. Canadian production did have a high in 2002 at 6.6 Tcf. However, the Canadian production has not dropped sharply as you claim; there is no evidence of a sharp peak (yet). Production in 2006 was still at 6.5 Tcf. Source: http://www.eia.doe.gov/emeu/international/gasproduction.html Kgrr (talk) 08:07, 24 December 2008 (UTC)

It depends on which numbers you quote, but Energy Information Administration (EIA) data indicates that US withdrawals did indeed peak at 24.1 Tcf in 1973, and then declined to 18.7 Tcf in 1983. However, after that dip they began to rise again and reached a new peak of 24.5 Tcf in 2007. Preliminary data indicates that 2008 production will be about 8% higher than 2007, mostly because of unconventional shale gas production. As far as Canadian production is concerned, Canadian regulatory authorities have been warning for some years that production is not backed by adequate reserves to sustain it in the long term. The National Energy Board (NEB) indicates in its latest study, Short-term Canadian Natural Gas Deliverability 2008-2010 that production appears to have plateaued in the 2000-2007 time frame, and its expectation is for a decline in production unless prices increase drastically. However, shale gas is the wild card in this deck. It has accounted for most of the recent US increase in production, and its potential in Canada has not been evaluated (although some people believe it is very high).RockyMtnGuy (talk) 16:20, 25 December 2008 (UTC)

The United States is now the reigning champ for the production of natural gas, surpassing Russia. U.S. Overtakes Russia as Biggest Natural Gas Producer. Crazy that this sort of think isn't reported more in the mainstream media. 76.105.219.46 (talk) 20:02, 13 January 2010 (UTC)pjc

"recent U.S. Peak predictions"

There's a line that says, "As of June 2008, the US Energy Information Administration projected US gas production to peak in 2022.[24] " My real question is, at what level (TFC/yr) does that predictions being cited, say U.S. gas production will reach? To research this main question I opened the .xls and see just production (line 18, which adds lines 19 and 20, while line20 is very small and (almost) constant) and "total supply" (line 26), neither of which are shown to peak in 2022. So my secondary question is for clarification (here, not in the main article) where are they reading that? My main one is, to please add the production level to the article, "...predicts U.s. production will peak at [LEVEL] in [YEAR]" format. Thanks.--Harel (talk) 00:16, 14 January 2010 (UTC)

I see that the latest update to the cited EIA spreadsheet was made December 2009. Per the latest version of the linked spreadsheet, the EIA now predicts US dry gas production to peak in 2009, decline slightly through 2013, then rise again at a slow steady rate through 2035. Plazak (talk) 01:13, 14 January 2010 (UTC)

Thank you Plazak for clarifying and for updating the entry itself. I've also calculated (23.27 / 20.6)^(1 / 26) = 1.00469846 and added briefly that this projection represents an average annual rate of increase of 0.47%/yr for 2009-2035. It's important that it be clear that "date of peak" is not the only relevant datum when thinking about peak, particularly if the level of the peak is (and thus the implicit avg annual rates of increase are)relatively low. --Harel (talk) 00:00, 16 January 2010 (UTC)

Another equation

this youtube video by Albert A. Bartlett has an equation at 4:36 which differs from others here. its called the "Expiration Time or "T sub E", of a non-renewable resource whose rate of consumption is growing steadily". I think its highly relevant to all of the articles on peak resource use and limits to growth. If anyone else thinks his equation is relevant, can they transcribe it? the math is beyond me, and i cant reconstruct it from the image on screen.Mercurywoodrose (talk) 02:07, 27 October 2011 (UTC)

Reversal of US decline

The US, far from experiencing "peak gas" is now producing the largest amount of natural gas in it's history. Ever.

http://www.eia.gov/dnav/ng/hist/n9050us2m.htm

This is not "Hubbert's Peak". This is not an "undulating plateau". This is a full on production explosion - a glut of gas.

Shouldn't this be incorporated into the US section of this article? The US has certainly experienced something that appears to repudiate the "peak gas" idea completely. 76.105.219.46 (talk) 01:12, 6 March 2012 (UTC)pjc

Could you elaborate? Gas is limited resource, so the production will peak inevitably. The only question is when. Few predictions turning out to be false doesn't affect the validity of peak gas.

As for the U.S. situation, the gas glut has been caused by high oil prices (surprising, isn't it?). This led to new, high cost, lower quality oil fields becoming profitable. These fields, incidentally, also produce a lot of associated gas, which flooded the market and depressed the prices. However, if the price of oil hasn't increased, these fields wouldn't have been drilled and we would see quite different picture in the gas market now. 1exec1 (talk) 23:42, 11 March 2012 (UTC)

The gas glut is related to high oil prices, but there is a lot more to it. There is a huge quantity of gas being drilled out of the Penn Marcellus, with very little associated oil. "Few predictions turning out to be false doesn't affect the validity of peak gas". Well if peak gas simply means "someday we will run out of gas" then teh whole article needs to be rewritten. The fact is that the current gas glut doesn't contradict "a few predictions" ... it contradicts all of them! Surely neutral POV would indicate this somehow, in the same way that neutral POV would point out the absurdity of James K. Glassmans 20K book. This is supposed to be a neutral POV page, not an Oil Drum page. 66.191.18.59 (talk) 19:08, 12 March 2012 (UTC)pjc

Can I ask you: how many %s does Marcellus shale represent in the total US production? 1exec1 (talk) 17:40, 18 March 2012 (UTC)

"Can I ask you: how many %s does Marcellus shale represent in the total US production?" Sure about 5% of total production. It was a lot less 5-10 years ago, so it represents about 15% of the increase from 2005. So it's clearly a big player in the gas glut we're experiencing. The broader point is that the exact opposite of the Hubberts peak has occurred for the largest gas producer in the world. If that doesn't discredit the "peak gas" concept, then you're peak gas belief is based in theory and not reality.(UTC)pjc — Preceding unsigned comment added by 76.105.219.46 (talk)

External links modified

Hello fellow Wikipedians,

I have just modified 8 external links on Peak gas. Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:

• Added archive https://web.archive.org/web/20080724113716/http://www.bu.edu/cees/people/faculty/cutler/articles/Natural_Gas_in_US.pdf to http://www.bu.edu/cees/people/faculty/cutler/articles/Natural_Gas_in_US.pdf
• Added archive https://web.archive.org/web/20080928150524/http://www.unctad.org/infocomm/anglais/gas/market.htm to http://www.unctad.org/infocomm/anglais/gas/market.htm
• Added archive https://web.archive.org/web/20081003081853/http://www.worldenergysource.com/articles/pdf/longwell_WE_v5n3.pdf to http://www.worldenergysource.com/articles/pdf/longwell_WE_v5n3.pdf
• Added archive https://web.archive.org/web/20100806145740/http://www.eia.doe.gov/pub/oil_gas/natural_gas/feature_articles/1997/intricate_puzzle_oil_gas_reserves_growth/pdf/m07fa.pdf to ftp://eia.doe.gov/pub/oil_gas/natural_gas/feature_articles/1997/intricate_puzzle_oil_gas_reserves_growth/pdf/m07fa.pdf
• Added archive https://web.archive.org/web/20080527233843/http://www.hubbertpeak.com/hubbert/1956/1956.pdf to http://www.hubbertpeak.com/hubbert/1956/1956.pdf
• Added archive https://web.archive.org/web/20090320104601/http://petroleum.berkeley.edu/papers/patzek/ArchivesofMiningPAS.pdf to http://petroleum.berkeley.edu/papers/patzek/ArchivesofMiningPAS.pdf
• Added archive https://web.archive.org/web/20081203021455/http://www.eia.doe.gov/oiaf/aeo/excel/aeotab_13.xls to http://www.eia.doe.gov/oiaf/aeo/excel/aeotab_13.xls
• Added archive https://web.archive.org/web/20131126081820/http://www.energywatchgroup.org/fileadmin/global/pdf/EWG-update2013_long_18_03_2013.pdf to http://www.energywatchgroup.org/fileadmin/global/pdf/EWG-update2013_long_18_03_2013.pdf

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

This message was posted before February 2018. After February 2018, "External links modified" talk page sections are no longer generated or monitored by InternetArchiveBot. No special action is required regarding these talk page notices, other than regular verification using the archive tool instructions below. Editors have permission to delete these "External links modified" talk page sections if they want to de-clutter talk pages, but see the RfC before doing mass systematic removals. This message is updated dynamically through the template {{source check}} (last update: 18 January 2022).

• If you have discovered URLs which were erroneously considered dead by the bot, you can report them with this tool.
• If you found an error with any archives or the URLs themselves, you can fix them with this tool.

Cheers.—InternetArchiveBot (Report bug) 01:06, 14 January 2018 (UTC)