The future of copper is uncertain!

Once the rate of oil production peaks and goes into decline, the rate of mineral extraction will likely fall off a cliff regardless of how much there is left in the ground.

Thanks, Jean!

I don't know if you are familiar with the new US Minerals Databrowser, produced by one of our readers, Jonathan Callahan, using USGS Data Series 140. This is one of the copper graphics from the Databrowser (which or course, does not disagree with what you are saying).

As I look at the graphic, and the others you produced, it seems to me that the graphs seem to suggest that historically, copper production has been governed mostly by demand, rather than availability. That is, production has ramped up and down to match need. That is consistent with your statements about cycles, and would make modeling difficult, I would think.

The recent run-up in price coincides with the run-up in oil price in the 2000s. It would seem like it might at least partly reflect the higher costs of mining, due to higher costs of oil inputs required for production. Some of the earlier reductions in cost may reflect improved methods of mining, and hence, lower needed price to cover costs and make a reasonable profit.

Its price went so high that copper cables are now often stolen, disturbing telephone and Internet communications.

This type of fact and the likelihood this trend will escalate in the coming years is often overlooked by those who profess that electrification has merit to "save" civilization.
On the local Detroit news, an article ran last week that 6' high chain link fence surrounding a ballpark was stolen. 100's of feet.
One wonders if it was stolen for the scrap value or to fortify some site.

Well, you don't burn copper. Most of it is still around.

With all due respect to Mr. Laherrère, I do not believe that the Hubbert Linearization technique has much to contribute when looking at copper or most other minerals. The real test of HL forecasts can be had by 'backcasting' -- only using data points up to some earlier date and then seeing how well the fit predicted production during the intervening years. Performing such validation tests will yield completely unsatisfactory results for the examples shown above. Another thing I find missing in most HL fits is any error analysis. My own investigations into using HL for oil production forecasts show huge error bars in most cases and leave me wondering why people continue to pursue this technique.

But HL is not the only focus of the article. I do believe that Peak ~Mineral~ is a very important issue and should be thoroughly investigated. Its is for this reason that we have set up the US Minerals Databrowser where one can generate graphs similar to those above for the entire set of minerals contained in the USGS Dataseries 140.

One of the charts in the US Minerals Databrowser addresses the prices from an economists' point of view by plotting adjusted price (in 1998 dollalrs) vs. global production. The assumption is that the commodity of interest is globally fungible with the US price a proxy for some general 'world price'. These charts for gold and copper help make sense of what Laherrère curiously describes as a 'chaotic' price evolution.

Having reviewd the price evolution charts for the entire list of minerals I can say that prices before WWII varied tremendously for many mineral commodities. Only after WWII did mechanization reach out to the furthest corners of the globe and bring rapidly increasing supplies to a genuinely international market. The price evolution charts therefore only cover the period from 1950 onward.

The general trend for most minerals since 1950 has been toward lower price and increased production -- toward the lower right in the charts below. Moves straight up as seen in gold circa 1980 indicate speculation while moves to the upper right indicate that a commodity is becoming increasingly dear, though not yet in short supply. In the two graphs below it looks like gold has reached an actual geologically driven peak -- increasing price at lower production, while copper is so far only becoming more expensive (dramatically so!).

It is important to remember that the interplay between supply and demand is also an interplay between geological constraints and economic necessity. The 'peak demand' meme is one to keep in mind. Browsing through the Minerals databrowser you will see that the asbestos world production curve should make it a poster child for the HL method as it fits the logisitic curve better than most other minerals. The only problem is that peak asbestos was determined entirely by a falloff in demand.

So what are the chance of demand for copper changing? Could we possibly get by using less copper if it were to become expensive enough? One more chart from the databrowser showing the 30 year history of US copper usage by industry demonstrates that consumption can go down in a pretty dramatic fashion as it did for copper between 2000 and 2003. For electrical products, the drop may be the result of offshoring the production of those products. But the housing industry was still strong in those years.

All in all, I think this is an important topic but I believe a modeling approach, particularly HL, is inappropriate. One gets a much better picture of the situation by actually reading the reports put out by the USGS. From the 2010 Minerals Yearbook section on copper:

World Resources:
Recent assessments of copper resources indicated 550 million tons of copper remaining in identified and undiscovered resources in the United States8 and 1.3 billion tons of copper in discovered, mined, and undiscovered resources in the Andes Mountains of South America.9 A preliminary assessment indicates that global land-based resources exceed 3 billion tons. Deep-sea nodules were estimated to contain 700 million tons of copper.

Substitutes:
Aluminum substitutes for copper in power cables, electrical equipment, automobile radiators, and cooling and refrigeration tube; titanium and steel are used in heat exchangers; optical fiber substitutes for copper in telecommunications applications; and plastics substitute for copper in water pipe, drain pipe, and plumbing fixtures.

Peak production and peak demand will be determined ultimately by peak price.

Best Regards,

-- Jon

I wonder if these graphs include still-expanding estimates for the Pebble Mine reserve in Alaska?
http://www.pebblepartnership.com/node/224

On Olympic Dam things are in limbo for now. The anti-nuclear State Labor Party was narrowly re-elected two weeks ago. The mine expansion needs huge amounts of electricity (.7GW), diesel and explosives. The water supply currently from groundwater needs to increase five fold via a desal plant on the coast 300km away. That desal seems unlikely to get approved with an announcement due in July. However under a Plan B the coast is still not spared from heavy activity. Most (70%) of the copper concentrate will be shipped to Guangzhou in China where both copper and uranium will be extracted.
http://www.world-nuclear-news.org/ENF-BHP_Billiton_outlines_Olympic_Dam_...
This helps in a number of ways; China has cheap labour and unrestricted coal power while it gets Australia off the hook with finding more water and energy. Shame about exporting all those jobs though.

I think this is the same person who said natural gas was about to peak and start a steep decline.
Those charts were very convincing as well

An interesting article, but along with others I don't think the techniques used for predicting conventional oil production can be used for predicting copper production.

Copper is much more akin to oil sands. The problem is not finding the reserves, it's figuring out a way to make money mining and refining them.

The decent (>10%) copper ores were mainly found and depleted by 1890 (http://www.theoildrum.com/files/JL_Cu_09_OreGrades.jpg). These were like the Spindletops of the mining world. Then the waterflood equivalents, the mines with >3% copper came into their own.

But by 1940 these too were just about depleted. Since then, the world had mined steadily lower grades of oil-sand-equivalents. As with oil sands, there are really large reserves out there, and the cut-off is not a specific copper concentration, but the economics of a particular deposit. The Olympic Dam mine in Australia, for example, is both economically more attractive because the associated uranium is very valuable, and politically dangerous as a result of being associated with that same uranium.

For those interested, at the ASPO-Denver conference this year, Vince Mathews, Director, Colorado Geological Survey gave an informative presentation titled World Mineral Demand and Security

Thank you for sharing this informative and insightful piece. I found the graph with the breakdown of copper's uses to be particularly interesting and somewhat surprising. When I picture the effects of peak copper I am prone to thinking about its effects on building and maintaining electric grids, especially for projects involving long transmission lines such as remote or offshore wind farms. However, I see that "electric and electronic" uses constitute only 19% of copper demand, while "building construction" demands the most at a whopping 51% of the total. I assume some of the total is structural material (bronze etc.) in addition to use in residential and commercial boilers and other domestic equipment. I'd also be curious to see a further breakdown of the building and construction component; does this include a building's internal electric grid, for example, or is that counted under "electric/electronic?"

It would also be interesting to see a chart illustrating changes in usage patterns over time, and I wonder whether an increase in the electric sector's share could be expected in a world shifting to dependence on wind, hydroelectric and other energy resources most available in increasingly remote areas, not to mention transporting power to or from increasingly remote unconventional oil mining/refining areas (see discussion of shale oil refining and Heading Out's post from this past weekend)? I see a parallel here to the increase in demand for natural gas from the power sector, a big part of the sustained price increases since the 1990's that have led to the development of unconventional gas reserves (shale), where the share of total gas demand from the electric power producing sector increased from around 20% to 40% over the period of a decade or so (I do not have the exact figures offhand but that is the general idea).

Here, the U.S. usage chart is from 2006, before at least the full-on collapsed housing and construction markets we have today. So it is also possible that an increase in demand from the electric power sector could be at least partly offset by a decline in demand for new construction uses, as well as by an increase in recycling.

Edits: Looking back up now, I see that one of the other commenters did provide this data above, and while drop in demand from the building sector does account for a substantial share of overall decrease in U.S. demand, demand from "electric and electronic products" has declined as well by an even greater amount. Granted development of renewables in the early 2000's was also slower than it is today, and I still wonder about overall share of demand from the electric grid, particularly since grid infrastructure needs upgrading in many areas regardless of use of renewables, and this area has not consumed a significant share of copper production in a long time. Substitution may occur first in other areas where the use of copper is less essential, such as piping and ductwork, or ornamental brass in construction.

To the poster above sarcastically (it seems) drawing the comparison to peak gas, a sustained rise in price (not seen yet) would be expected along with a production peak, but a rise in energy costs and progressive decline in ore quality and location, and accompanying increase in environmental and other externalities, suggest that price increase could be on its way. You could say that with production of many metals shifting to large-scale labor-intensive and environmentally destructive mining in the Third World, as well as the emergence of black markets in some cases such as rare earths, we have already seen an increase in the human toil (and bloodshed) cost of metals, even if this has not translated into a price increase in the "civilized" world.

And even if North American gas reserves have dramatically increased and the price leveled in recent years, we have in fact seen the long-term sustained price increase mentioned above that has made unconventional gas economically (though not necessarily energetically) viable. Again, what remains to be seen is whether long-term supply predictions pan out or whether they sputter due to poor EROI of hydrofracking and other energy intensive exploration and drilling activities, or environmental problems like air and water quality or CO2 intensity. Keep in mind we are only a few years into intensive shale gas production on this continent, but already it has reached 40% of total production in the U.S.; meanwhile, it has become clear already that the activity is far from the innocuous and noninvasive job creation machine many gas companies initially claimed when staking out new drilling territory in the Marcellus and elsewhere. The same could be said for energy-intensive copper mining operations with large footprints.

The questions that spring to mind:

1) how much of copper's current uses have an easy substitute?

US pennies went to zinc about 25 years ago;
Copper pots have gone to aluminium and stainless steel;
Most electrical wiring can go to aluminium, as long-distance transmission has;
Pipes can go to plastic, iron, etc;
Telecom wiring can go to fiber optic;
Wood treatment can go to zinc?

What's left? Internal electronics wiring/circuit boards?

What are the %'s for the above? and

2) how much can be recycled, and how much is recycled now?

Anybody mentioned Mongolia?

Today's news:

Rio Tinto has finalised a new agreement with the Mongolian government for the development of a $4.4 billion copper-gold project in the country.

Under the investment agreement, the Mongolia government will own 34 per cent of the company holding the license for project, which is being developed by Rio Tinto and Canada's Ivanhoe Mines.

Rio Tinto says the development phase of the project will now start, with production to commence in 2013.

Output will be ramped up over five years to a full production rate of 450,000 tonnes of copper per year, with substantial gold by-products.

The $4.4 billion project was initially expected to cost closer to $2 billion, and Ivanhoe originally planned to commission the mine last year.

source: http://metalsplace.com/news/articles/33565/rio-tinto-in-a-44b-copper-gol...

And then there's Snowdonia, North Wales, but we don't allow mining there because it nice countryside (seems we don't care about wrecking the countryside elsewhere as long as it's a long way away).

The (full, un-ellipsissed) link to the Gavin Mudd paper is:

http://civil.eng.monash.edu.au/about/staff/muddpersonal/2009-CMS-03-Base-Metals.pdf

Full Reference:
Mudd, G M, 2009, Historical Trends in Base Metal Mining: Backcasting to Understand the Sustainability of Mining. Proc. “48th Annual Conference of Metallurgists”, Canadian Metallurgical Society, Sudbury, Ontario, Canada, August 2009.