Thoughts from the Edge

Many people feel we are now on the edge of something new. The edge of what is hard to discern but its there. Perhaps something good perhaps bad and probably both. Thoughts from the edge looks back at the path we have taken and considers what lies beyond.

Thursday, February 25, 2010


Before peak oil there was this saying that the stone age did not end because they ran out of stones after peak oil the saying was they did not not start making stone arrow heads again because they never ran out of glass and steel.

This Week In Petroleum => This Week In Propaganda

Watch the forest see the trees

Complex system analysis requires and understanding of the whole without losing sight of the individual components.

A dead man does not use gasoline or food.

No wheels no roads, no roads no armies.

Saturday, February 20, 2010

Checking The Exhaust Pipe For Peak Oil Determining Oil Consumption From Atmospheric C02 Measurements

In Chemical Engineering the concept of mass balance plays an important role in the design of chemical plants. This simply means that the total mass of the products and reactants are constant. If you burn a fossil fuel then the mass of oxygen and fuel equals the resulting water and carbon dioxide.

Thus, all the fossil fuels that have ever been burned are redistributed as C02 into the atmosphere and eventually into other sinks. Since its balanced this complex flow can be modeled with very simple equations. Thus we can treat the world as one giant chemical factory converting fossil fuels to carbon dioxide. The shear size of this global chemical factory ensures that linear changes are a decent approximation of the balance between C02 and fossil fuel inputs on a short time scale. Of course the total amount of C02 is steadily increasing thats not the assumption made its that the flux of C02 into the atmosphere is linearly depending on its inputs.

Oil production is determined by assuming a linear relationship between anthropogenic C02 sources, Oil, Coal, Natural Gas and an estimate of non fossil sources and atmospheric C02 levels. This relationship shows and excellent fit with oil prices and shows that peak oil production is well in the past.
This is not however a paper about global warming, lets save that for a future discussion.

As I have learned more and more about peak oil my concern has moved away from the reasonable assumption that extracting a resource that takes millions of years to form at a furious rate is a problem to realizing that the data available on the subject is suspect and politically sensitive. What data we do have is provided by oil companies that either claim peak production is an event of the distant future or that its not even possible despite the basic order of magnitude mismatch between the rate of oil formation and current rate of consumption. Deducing oil production from C02 production offers a novel approach to checking the production data provided by the oil industry.

The realization that this might be possible dawned on me as I was looking at the graph of C02 measurements from Mauna Loa Observatory.
Figure 1.

I noticed a flattening in the late 70's near peak US production and other slight dips or flattening that corresponded well with US recessions.

Courtesy of CalculatedRisk here is a graph (Figure 2) over a similar time range I don't have the data to merge the two graphs but note that economic activity can slow substantially before an official recession is recognized or it can be sudden. The CO2 graph is arguably a better barometer for world recessions than economic data.

Figure 2:

In (Figure 3) a graph of all the input data is presented.
Mount Pinatabuo erupted in 1991 and may have influenced world wide C02 levels both via venting and cooling effects however it also happened to coincide with a recession so its difficult to distinguish between the two cases for that year. Fate somehow ensured that the one natural event that could perturb world C02 levels overlapped with a recession.

Figure 3:

Outside of economic data another factor effecting C02 emissions is population growth. In (Figure 3). World population estimates are taken from:

Not surprisingly the relationship is linear however the slopes of the two lines differ given that C02 usage is associated with wealth. For example, 300 million people in the US use 25% of our carbon resource base. One could easily deduce that as the population has grown the per capita wealth has declined. Of course, we know this is true from numerous sources so its not an earth shattering deduction however already we have gleaned some valuable information. As we go further the most important fact is that the relationship between C02 emissions and population is a simple linear function. This is the fundamental fact that we can exploit to deduce real oil consumption using a linear relationship.

Next we need a bit more information given we have a simple linear mapping to population if we know overall Coal, Natural Gas and Oil consumption then it makes sense that the relationship of these to C02 must also be linear. Coal, NG and reported Oil production are in terms of million tone of oil equivalents (mtoe) using the BP data and converted into to C02 using their conversion factors.

C02*(linear function) = Coal + NG + Oil+ ( None fossil emissions)*population

This is an equation that can be solved.

To solve we limit our C02 data set to the consistent Mount Loa observations although with work it could be extended back in time. They are complete from 1965-2008. Next the BP Statistical review of world energy contains a nice breakdown of world energy usage in normalized units in terms of coal, Natural Gas and oil. Along with standardized conversions to C02 emissions in the BP spreadsheet.

To include the population effect non fossil fuel emissions are assumed to be 1 metric ( tone )ton or tonne? of C02 per year however given the relationship is linear thus actual value of this assumption is not critical. The choice of 1 however was not random, but based on a search of the available literature with this result in (Figure 4) chosen as a decent estimate.

Figure 4
The choice of 1 is slightly above the low income average and I feel very reasonable Also it simplifies the math.

Now given the historical consumption of Coal, Natural Gas and Oil we simply solve for the linear relationship with C02 emissions making the assumption that neither Coal nor Natural Gas production have been politically altered and that changes in Oil production numbers vary with the political climate. The solution should be a perturbation of the assumed linear equation over the time span considered. Thus, a simple least squares fit is all that is required to remove any politically induced variation in reported oil production. The published data is used as input to first solve for the assumed linear relationship.

(linear function) =

(Coal + NG + Oil+ ( None fossil emissions)*population)/CO2ppm.
The linear function is assumed to be the least squares fit of the published data. This basically assumes that error in the data particularly oil are not systematic and probably within 5%-10% over a multi-decade time span. Unit conversion is also buried in the constant. Perhaps it can be considered to be a response factor to rising CO2 levels although a physical interpretation of the value is not made. It would be some sort of real physical constant. This is effectively a sort of mass balance equation and is the model for our earth scale chemical factory.

Figure 5:

Notice the fit is excellent during the period of stable prices during the late 1980's to 1990's. Also notice it looks like oil production was under reported during the portion of the era preceding the OPEC embargoes and peaking of US production in the 1970's. This is not surprising since decent accounting of global oil production in the late 1960's was not exactly a high priority. Deviations from a linear assumption already fit well with what we know about history. The correlation of the discrepancies between our linear assumption and historical events regarding oil already hint that oil production numbers might have a political bias.

Now at this point we finally have enough information to calculate real oil production using our assumptions. We simply use this linear fit to solve for oil production.
Oil = C02ppm*(linear function) – Coal – NG.
This is compared with published oil production and non inflation adjusted prices and inflation adjusted prices as a check.

Figure 6:

This is the key graph of this report. The price is using the non-inflation adjusted or real dollar price of oil. At the time this better illustrated the percentage change in price from oil production alone. Associated monetary inflation related to oil production can cause distortions. I find the relationship between price and my calculated oil consumption simply stunning !

Next we include inflation adjusted dollars, obviously the way we adjust for inflation has issues but still price is strongly correlated with my calculated oil production. The differences of course suggest a lot about how our financial system really works. If one included the housing and related credit bubbles in their inflation definition then I doubt we could even fit our inflation adjusted graph without resorting to a semi-log plot. Its not unreasonable to suggest our precarious financial state is the result of our leaders response to falling oil production. We have every indication that the next step will be that common one when a gambling addict has lost massively on a gamble left unchecked it looks like we will double down. The final outcome is fairly certain.

First its rather obvious that the great drop in oil consumption during the 1980's was a myth that never happened. In fact the Arab oil embargo during the 1970's barely shows on the graph. What's more important however is the 1980's since it's the model most people use to claim oil consumption will drop sharply with price. It's clear that this never happened. What's important is we never did reduce oil consumption in response to rising prices instead all that happened is oil consumption flat lined as population increased back in the 1980's. Of course just as obvious in the last ten years despite the rapid increase in prices oil production fell linearly. Next the period of above reported production during the 1990's fits very well with technical advances in particular horizontal drilling. The super straw effect is clear and obvious.

Looking both backwards at the 1980's and forwards its clear that current fairly low oil prices are probably a temporary phenomena related to short term financial dislocations and probably other approaches to keep the system going. Looking forward to 2009 and later population pressure is relentless and if you look at refinery utilization rates and oil tanker rates it's obvious that the supply side has seen a dramatic drop during late 2008-2009.

Next lets look at the flux the total amounts where used for a reason since if there was no absorption phenomena the total amount of C02 in the atmosphere equals the amount produced if we have identified all sources the mysterious fit constant hides the complexity of absorption processes.  Theoretically the rate of absorption should increase as the production rate increases thus the linearization of the total and the derivative should be parallel.

Whats interesting is that the rate of C02 increase is not keeping up with the input flux. In other words a buffer is absorbing the C02 in proportion to its rate of increase this of course is a classic chemical buffer effect where is long as the buffering agent is in excess it can absorb the inputs. The obvious buffer is the oceans. So AGW is a bit of a myth the problem however is for how long since the response of a buffer is highly non-linear when it reaches saturation. This is not I repeat a paper on AGW its about C02 however this divergence suggests that a buffer of unkown capacity is being saturated and once it is ...

Back to the present lets look at 2008-2009.

Figure 9:

Whats most interesting is it seems that perhaps oil was deliberately removed from the market leading up to the price spike in mid 2008. My opinion is since OPEC could not control prices by pumping flat out they chose a different route. Deliberately store oil even as it leads to a spike dampening consumption then flood the market thus it suggests most of the oil that later seemed to flood the market as the economy was crashing was perhaps deliberately stored earlier despite its effect on prices.

Desperate yes but it explains a lot.

We can see more weakness or dips from the economic crash and the hurricanes at the end of 2008 followed by a spike at the beginning of 2009 when oil reached a low in price. Its not clear if anything more then a qualitative view is possible at this level of detail. However at least a short term glut that was able to flood the oil distribution network seems very possible. A few hundred million barrels is more than enough to flood the system especially considering the economic panic. It seems that the slope of increasing C02 levels might have changed from prior years and flattened somewhat.

However, given the broad nature of the collapse, coal and natural gas consumption were almost certainly affected so this decline cannot be attributed to oil alone. It looks like 2008-2009 may well have been the year of peak overall fossil fuel usage. There is a good chance that a significant increase in the natural decline in overall oil production also started at some point in 2008 its impossible to know for sure. However if so it makes sense that this perhaps could have triggered the series of interesting events of 2008-2009. Assuming a significant quantity of oil was stored first at high prices then perhaps later as economic turmoil and hurricanes cause the global economy to grind to a halt then as this oil is consumed without a sharp increase in C02 means the underlying daily production is significantly lower. The C02 data at least qualitatively supports the possibility of both significant quantities of oil being stored and overall significantly lower daily production. Even if true we can't know if its a deliberate cut in production because of falling demand or a geologic decline. At least so far however it seems reasonable to guess that what really happened in 2008-2009 will eventually become clear in 2010 as stored oil is burned and daily production has to rise to make up the difference.

This makes one wonder about the renewed interest in trading of carbon emission credits. Making money not burning fuel you don't have is a pretty good deal. This does not mean coal or NG reached their geologic peak but the constraint of falling oil production limits any further expansion in the exploitation of these two resources. Even if for example CTL or NGL process eventually become viable overall economic shrinkage between now and then will probably result in at best a lower peak in production. Expansion of coal consumption at anything close to the rate of China on a global scale is probably impossible. Regardless, evidence for a fairly large if short term glut in oil availability from the rapid rate of the economic collapse seems very plausible.

Also note that the shift from oil to coal has an impact on the CO2/SO2 ratios as the increasing levels of S02 emissions can cause weather changes. So although I do not want to discuss global warming there is a potential short term weather effect. We should see slightly colder temperatures and more precipitation as Chinese coal production ramped up similar to volcanic emissions. I did not find good global numbers for S02 concentrations, but if measurements rather than estimates from fossil fuel production can be found they can act as a check on coal production.

In conclusion the thesis that a linear relationship between fossil fuel consumption and C02 levels in the atmosphere seems valid when checked against price which is generally independent of manipulation over the period studied. Significant differences are found between reported production levels and calculated levels during periods of high prices while the curves are in much better agreement when production is growing and price is falling. The financial bubbles actually correspond well with falling oil production and rising prices. Not only does it seem our political leaders are attempting to hide the real situation in oil but their solution to the problem seems to be to blow ever more damaging financial bubbles. I think the people of the world should have some say in our future and its clear that our leaders have abused their position of public trust in a futile attempt to maintain the status quo.

Global Warming:

I don't want to say to much about global warming however these results are alarming we are living on borrowed time if you will in a number of ways. First and foremost they suggest that global warming i.e acutal temperature increases should be very mild over the time period studied. With the strong increase in unscrubbed coal emissions the short term cooling effect of S02 should domininate. Next there is obviously a very large buffer involved and this is probably the Oceans so they will act to offset C02 emmisions. The problem is these effects work until the do not in both cases highly non-linear thresholds are reached. So the good news is AGW is probably very weak at the moment the bad news is that as the effects making it weak are short term the final effects are far larger than can be anticipated. They are large enough that they cannot be modeled without and excellent model. In short they are off the charts.
The relationship between C02 flux and temperature is and explosive system changing exponentially. The system should change radically in a period measured in years if not months. These spikes are simply not captured in the historical record as we lack the technology to capture them. However the follow on massive extinction events are well documented. Small wonder meteor strikes are picked as a culprit. The problem is this spike phenomena is intrinsic.  The trigger for follow on extinctions is short and devastating and built into the system. Thats all I want to say about global warming but I suspect its sufficient.

Resources and Links

Mount Loa C02 data

BP Statistical Review

I found this link that uses a very similar approach and different conclusions.
But it also highlights that the C02 data and claimed consumption either don't match or the planets response is non linear. Regardless of interpretation of the data he also gets the divergence.