The human machine and the internal combustion engine are surprisingly similar, despite outward appearances. They both operate at a thermodynamic efficiency of ~30%, the former oxidising primarily carbohydrates, the latter oxidising hydro-carbons to release chemical energy (MacKay, 2009).
While most cars rely on explosions and associated high temperatures, muscles operate at 37 degrees celsius. Still, the carbon based chemical inputs and outputs (water vapour and CO2 in both cases) are remarkably similar. In fact, Odum (1971) describes the “cold fire of the body's metabolism”. Howard Odum put a lot of effort into visualising this equivalence with his energy diagrams:
Given these similarities, it is surprising that so little is known about the relative efficiencies of humans and industrially produced motors. We know that human muscles can be applied much more intelligently to a complex task (e.g. compare the art of scything with the brute force approach or a petrol powered lawn mower). Nowhere is this more controversial than in the realm of personal transport, where conflicting claims are rife. Some of these claims are sweeping and without context (e.g. “eating meat is worse than driving a 4 by 4”) and few of them have been subjected to the rigors of peer review. Scientists, in general, tend to focus on more high-tech issues, although arguably the human body is the most highly advanced machine we know about.
In this article I will put the best available numbers on the CO2 emissions from human respiration on the table and subsequently speculate about the implications for estimates of the CO2 emissions associated with cycling.
Evidence suggests that a typical human body at rest burns food at a rate of around 80 W. The table below, from Webster et al's (1985) callorimiter tests, shows this. Non-physical work does not raise the average heat loss by much.
This information allows us to calculate the proportion of Total Primary Energy Supply (TPES) that is consumed by humans: 15 TW on average according to BP - Vaclav Smil puts the values at closer to 12 TW, but let's take the larger number to be conservative of human power use. Let us assume there are 7 billion humans each using an average of 50 W (lower than the 80 W figure to account for sleeping and that not everyone is a healthy adult). The proportion of TPES used by these people would be calculated as follows:
(p.human <- 50 * 7 * 10^9)
## [1] 3.5e+11
(p.humanity <- 12 * 10^12)
## [1] 1.2e+13
p.human/p.humanity
## [1] 0.02917
The result is 3%, a surprisingly high result considering that we have yet to factor in the energy costs of food production. If we assume that the average carbon intensity of human metabolism is similar to that of average power supply, the CO2 emissions of human respiration would be around 3% of anthropogenic emssions also. But is this so?
Nolt (2011) provides a convincing method to estimate the average CO2 emissions from breathing per person per year as 0.2 tonnes. This is equivalent to 547.9452 gCO2 per day. (This is roughly 1/3 less than one internet source reported by Steve White, personal communication.) Without delving into the methodology, it is clear that, if true, the lower figure would represent 1.4 Gt of CO2 breathed out per year:
0.2 * 7 * 10^9
## [1] 1.4e+09
Global anthropogenic emissions in 2012 were around 35 GtCO2 (as reported on the BBC and elsewhere) - implying that humans directly breath out 4% of our total emissions:
1.4/35
## [1] 0.04
The carbon dioxide emissions of cycling have been estimated by Coley et al (2002) to be 8.3 g carbon per kilometre, including the embodied energy costs of food production. Knowing that the food:production ratio 1:5.75 was used in this study, we can work back to estimate the CO2 emissions of respiration whilst cycling as follows:
(co2.km.Coley <- 8.3 * 6.7)
## [1] 55.61
co2.km.Coley/5.75
## [1] 9.671
The result, that cyclists breath out CO2 at a rate of 10 g/km suggests that cycling is around 15 times less carbon intensive than driving, assuming single occupancy cars and ignoring all wider boundary impacts.
This 10 g/km figure is substantially less than the CO2 emissions of respiration estimated in a back-of-the-envelope calculation by Steve White (personal communication). The discrepancy arises, it seems due to White's assumption of a doubling in CO2 output (metabolic rate) during cycling. In fact, Metabolic Equivalent values can easily exceed 4 for moderate activity, far more than the value of 2 assumed by Steve White.
Of course, the overall environmental impacts of cycling are much more complex, and relate to many interrelated factors from the demand for cars and tarmac to propensity to shop locally. For discussion of these wider rangin energy impacts, the reader is directed to my first paper on the subject (Lovelace, 2011).
This is far from the last word on the matter and much more research into the energy and emissions savings from cycling is needed.
Lovelace, R., Beck, S. B. M. B. M., Watson, M., & Wild, A. (2011). Assessing the energy implications of replacing car trips with bicycle trips in Sheffield, UK. Energy Policy, 39(4), 2075–2087. doi:10.1016/j.enpol.2011.01.051
Mackay, D. J. C., Energy, S., Air, H., & Cam-, U. I. T. (2009). A plan with a time-line. Energy.
Nolt, J. (2011). How harmful are the average American’s greenhouse gas emissions? Ethics, Policy and Environment. Retrieved from http://www.tandfonline.com/doi/abs/10.1080/21550085.2011.561584
Odum, H. T. (1971). Environment, power, and society. Wiley-Interscience.
Webster, J., Welsh, G., Pacy, P., & Garrow, J. (1986). Description of a human direct calorimeter, with a note on the energy cost of clerical work. Br J Nutr, 1–6. Retrieved from http://journals.cambridge.org/production/action/cjoGetFulltext?fulltextid=858672