We’ll keep charging forward with the discussion about the value of human labor in manufacturing and how to consider it in life cycle analyses of differing processes and techniques of production (see part I here).
First, a clarification. This posting has nothing to do with forced labor, illegal child labor, excessive conditions and workplace violations, improper safety standards for labor, low wages, etc. These are all critically important issues potentially affecting a wide range of companies and their manufacturing procedures – but, that is not what this discussion is about.
Our concern here is how to respond to this question: if a company reduces the amount of machinery used in manufacturing and replaces that machinery with manual labor, does that help from an environmental or green manufacturing perspective? I postulated that for assembly tasks one might make the argument that more human labor (replacing automation) might produce the product using less energy and resources and, ultimately, making the product easier to disassemble at its end of life. But we’d need to consider the quality of the labor (meaning is it dull and repetitive or intellectually stimulating and, for sure, is it free from danger or other safety issues?).
It is conceivable that understanding this “tradeoff” might define a new economic situation that could actually encourage improved workplace environments for manual workers. We can see.
But, for the meantime we’ll focus on the analytical basis for answering the above question. And our discussion in this posting will be rather academic.
To recap, the methodology being described here from the paper we wrote and referred to in the last posting is related to economic input-output (EIO) LCA. Energy of labor and EIO-LCA should not be applied at the same level of analysis because many sources of energy use would be double counted. However, energy of labor can be very effective if incorporated into hybrid process-based EIO-LCA. The energy use of labor enriches the horizontal scope of process-based LCA, while EIO captures vertical supply chain impacts.
The energy use of labor helps address the disparities between environmental and economic accounting. Environmental analysis largely ignores labor, while the cost of labor factors very heavily into economic analysis. Evaluating the energy use of labor can help reduce the gap between those who prioritize environment and those who prioritize economics.
Finally, human capital, like environmental capital, has externalities that can be passed from a manufacturing system to society at large. For example, manufacturers who pay workers less than a livable wage rely on social programs to support their workforce. The energy use of labor is a tool with which we can begin to account for the environmental externalities of labor.
How do we estimate the energy use per worker-hour?
Three straightforward methods of estimating energy use per worker-hour (EPWH) are presented to produce a lower bound, an upper bound, and a value appropriate for use in life-cycle assessment. The methods are respectively derived from human metabolic activity, total primary energy supply, and non-industrial energy supply and described below.
Metabolic activity – A lower bound estimate (one that’s likely to give the lowest estimate) of energy use per worker-hour is given by human metabolic activity. An active individual can expend 2800 kilocalories per day or, on average, 0.5 MJ per hour. However, this method fails to consider the much greater amount of energy embodied in and used in the infrastructure employed to support labor. Nor does this consider efficiency losses from food production to digestion.
Primary energy supply – An upper bound estimate (that’s one that is likely to give the highest estimate) is given by amortizing a country or region’s energy supply across its worker population and over the number of hours in a year.
Last time we referred to the work of Odum. In his book “Environmental Accounting (Odum, H. T., 1996, Environmental Accounting: Emergy and Environmental Decision Making, Wiley and Sons, New York) he calculates the national fuel share per person based on the general population. Based on 1993 data, he concluded that 967 MJ are expended per capita per day or approximately 40 MJ per capita per hour.
However, not all members of the general population are productive workers at any given time. Just as a machine tool must be manufactured and have an end of life, a worker must have a childhood and an end of life. By amortizing energy use over the worker population, we account for the full life cycle of the worker. We therefore allocate energy use over the worker population, as opposed to the general population, to give us a better estimate of the contribution of labor to the energy use of a production system.
This upper bound estimate considers all the infrastructure and services that go into supporting a worker in terms of primary energy. Primary energy is measured in the units of tons of oil equivalent (TOE). Unlike final consumption in the form of refined fuels or electricity, primary energy captures all transformation and distribution losses.
However, energy use per worker-hour (EPWH) calculated based on primary energy cannot be used as a component of process- based life-cycle assessment because this method double counts industrial energy use.
Non-industrial energy supply – A better estimate of energy use per worker-hour for the industrial sector is derived from non-industrial energy supply, which includes all primary energy except that supplied to industry, as given by the equation below:
EPWH = (TPES – IPES)/(population x hours / year)
where TPES is a country or region’s total primary energy supply and IPES is industrial primary energy supply. IPES can be replaced with primary energy supply to other sectors of the economy or specific industrial sectors, such as the petrochemical sector, to reflect a particular product or process.
Energy use per worker-hour, in terms of primary energy, captures the energy mix and efficiencies in transformation and distribution for a given region. However, IPES is not always readily available, so we approximate it using industrial final consumption (IFC) and total final consumption (TFC) of energy calculated as follows
IPES = TPES x IFC/TFC
This assumes the ratio of final consumption to primary energy supply for industry is representative of the ratio of final consumption to primary energy supply for the country. Countries with industries that consume disproportionately more primary energy than the country at large are penalized by this assumption, resulting in a larger value of EPWH.
The International Energy Agency (IEA) regularly compiles and publishes values for TPES, IFC, and TFC from each country or region in its purview (see IEA website). As defined by the IEA, the industrial sector includes mining, smelting and construction but does not include transportation used by industry. The most current data available reflects 2004 activity.
The International Labour Organization (ILO) is a branch of the United Nations that similarly compiles employment statistics on an annual basis. Worker populations include civilian workers over an employment age, which is typically 14-16 years of age. Though there are disparities in what each country reports, data from the IEA and the ILO is likely more reliable than data compiled from each country directly.
Of the three methods discussed, amortizing non-industrial energy supply (the last method presented above) yields the most accurate estimate of energy use per industrial worker-hour for use in process-based or hybrid economic input-output life-cycle assessment. We’ll use this method for the examples and accompanying discussion in the remainder of this presentation on energy of labor.
We’ll pick up from here in Part 3 next time with an example comparing the differences between workers and manufacturing machinery.
David Dornfeld is the Will C. Hall Family Chair in Engineering in Mechanical Engineering at University of California Berkeley. He leads the Laboratory for Manufacturing and Sustainability (LMAS), and he writes the Green Manufacturing blog.