Summary.    This article describes a project undertaken by the Japanese company Hitachi Energy and three of its suppliers in conjunction with the E-Liability Institute to understand where CO2 emissions were produced in the value chain for the copper used to manufacture its...

Companies are increasingly facing pressure from stakeholders, external and internal, to decarbonize their operations and mitigate their impact on climate change. But without a rigorous global system for carbon accounting, such pressure can feel more rhetorical than substantive, as companies are, at times, goaded to “go green,” become “net zero,” or embrace “the circular economy.” These terms have no well-established definitions, so they can become veils for stakeholders to drive ideological agendas unrelated to decarbonization or for companies themselves to engage in greenwashing.

One way to forestall such distractions and to focus companies on meaningful decarbonization is through rigorous carbon accounting that operates on at least the same scale of accuracy and verifiability as financial accounting. In a 2021 article in Harvard Business Review, one of us (Ramanna), together with Robert S. Kaplan, introduced just such a system, called E-liability, and since then, we have piloted its use in several major organizations through the not-for-profit E-liability Institute.

Recently, Hitachi Energy approached the institute to join the ranks of companies using E-liability principles to accurately measure emissions in its production, purchasing, and product designs. Hitachi had a specific problem it wanted to address, and it found the E-liability approach could be useful in generating better data for a solution.

The company had been under pressure to use recycled copper in its manufacture of electric transformers, instead of virgin-mined copper. The pressure came from those who felt that recycled inputs are “greener” and therefore better for the planet. But engineers and managers at Hitachi familiar with the intricacies of transformer supply-chains were skeptical, and they needed an accounting method to help them come to a data-driven answer.

Our pilot with Hitachi focused on isolating the virgin-copper supply chain for electrical transformers, working with representative suppliers of copper wire, their suppliers, and eventually the mining company at the beginning of that value chain. In parallel, the pilot explored the emissions associated with sourcing and producing transformers using recycled copper, allowing for a direct comparison between the two alternate inputs. While the pilot did rely in places on approximations of emissions rather than measured outputs, it nonetheless generated a template for how Hitachi and its suppliers could accurately calculate emissions across different production techniques.

Structuring the Pilot

As a first step, the company identified one plant across its own production sites for the study: its transformer factory in Brilon, Germany. The general manager of this plant joined one executive each from Hitachi Energy’s central procurement and sustainability teams to serve as the Hitachi point-persons on the pilot.

Next, the company had to line up its suppliers for the copper used in electrical transformers at Brilon, all the way up to the copper mine. For the purposes of the study, the company created a fictionalized supply chain three tiers deep, involving firstly, Hitachi’s direct supplier Dahrén, which enamels the copper wire used in its transformers; secondly, Dahrén’s supplier Elcowire, which casts copper wire rods from copper cathodes; and lastly, Elcowire’s supplier Boliden, which fabricates copper cathodes from both virgin-mined and recycled copper.

Dahrén, Elcowire, and Boliden are all Swedish-based companies, and Boliden’s copper mines are also located in that country. One project manager (all supply-chain professionals) from each company was designated to join the pilot team, bringing the total team strength, besides us, to six.

The third step was for the project team to agree to simplified production diagrams for the value chain depicting the processing stages of copper in transformer manufacturing. To keep the study tractable and achievable within a short period, the team employed the Willie Sutton Rule, named for the famous bank robber who explained his choice of profession as “focusing on where the money is.” Adapting this rule to emissions, the pilot team identified key high-emissions processes in the value chain, thereby abstracting away from potentially hundreds of other processes. The focus of the pilot was on learning how to do proper carbon accounting rather than on trying to account for everything all at once.

Finally, the individual companies went about calculating the product-level carbon in their respective outputs to their immediate customers. For simplicity, the pilot focused on making these calculations over a monthly cycle – that is, the emissions for various input activities were calculated for a given month, to be then allocated to outputs produced in that month. The objective was to do this in a way that could be continually replicated across successive periods (and batches), so that subsequent iterations could improve the accuracy of the product-level emissions numbers generated.

The Results

The value chain begins with Boliden, a primary-sector company that sources copper in three different ways: from third-party mines overseas; from their own “low-carbon” mines in Sweden, and by recycling copper from a variety of sources. For the purposes of this pilot, the recycled copper was assumed to be harvested from electronic waste such as junked circuit boards, a recycling source chosen partly because the associated process is emissions intensive. Boliden’s refined copper cathodes sold to Elcowire are thus of three types, depicted in the below exhibit as Products 1, 2, and 3.

The three products do not differ in any other way than in the sourcing of the copper, because copper from the three different sources is fungible in use. In fact, Boliden’s interest in this pilot was spurred by its own low-carbon copper product: trying to sell a commodity like copper as a differentiated low-emissions material requires a credible cradle-to-gate accounting system like E-liability.

For the purposes of calculating its emissions to be allocated to these three outputs over a given month, Boliden set out by defining the key relevant activities that are the sources of emissions. These are represented in the above exhibit as “Purchases,” “Mining,” “Recycling,” and “Refining.”

The first three activities each relate exclusively to one product – thus emissions “purchased” are those embedded in the raw copper acquired from overseas mines, and such emissions need only be allocated to Product 1. Likewise, the emissions associated with mining are those, for a given month, from operations like blasting, trucking, and ore concentration that are only relevant to Product 2. Thus, the emissions associated with the first three activities can be directly allocated to the three products, based simply on the unit usage of these activities per ton of copper output.

The fourth activity – refining – applies to all three outputs, and so, the refining-related emissions for each month are allocated to all of Products 1-3 in that month. In this case, the refining hours per unit output happen to be similar across the three products, allowing for a simple proportional allocation. (You can find the presentation of a more detailed version of Exhibit 2 here.)

The end-result of this allocation at Boliden shows the different emissions profiles across its three products, calculated as CO2 per ton of copper cathode for the given month. Although the numbers in this exercise are stylized and should not be relied upon as indicative of the true emissions numbers, they are likely directionally correct. In effect, they show that Product 1 (copper from an average overseas mine) and Product 3 (copper from recycled electronics sources) are roughly similar in their embedded emissions (3,730 kg and 4,516 kg of CO2 per ton of copper cathode, respectively). Product 2 (copper from Boliden’s own mines) has substantially less embedded emissions (1,240 kg of CO2 per ton of copper cathode). Boliden attributes this result to the high productivity of its Swedish mines (relative to many overseas mines) and to various emissions-savings activities, such as having a high rate of electrification within its mines and drawing from a low-emissions grid.

Product 3 (recycled copper) has a high emissions profile because the recycling process involves incinerating plastic circuit boards to harvest embedded copper, releasing large quantities of GHGs into the atmosphere. The implication of this analysis for downstream companies in value chains is, of course, not to stop using recycled copper, but to work with suppliers to ensure that cleaner methods of copper recycling can be developed. This may even include working with circuit-board manufacturers to use low-emissions plastics in their production, so that the copper can be extracted for a second life in a way that is both economically and environmentally sensible.

As the copper moves downstream, the emissions added reduce sharply, as shown in the second exhibit. The main sources of emissions at Elcowire, Dahrén, and Hitachi are (i) energy consumption and (ii) the allocation to products of the emissions embedded in the companies’ physical plants (buildings, equipment, etc.), a process akin to depreciation. Put differently, Elcowire, Dahrén, and Hitachi Energy’s transformer operations appear to be low carbon-margin businesses, with comparatively little ability to influence the carbon-efficiency of their own outputs, apart from making different purchasing and energy-sourcing decisions.   Of course, our pilot study excluded many direct sources of emissions, and a full counting and allocation of all such emissions could reveal new opportunities decarbonization and competitive differentiation in these seemingly low carbon-margin businesses.

An interesting feature of this supply chain is that Boliden is in fact a customer for Hitachi’s transformers, which it uses in its own mines.  This means that in buying a transformer Boliden also acquires responsibility for the emissions embedded in it.  In a second round of allocating emissions, therefore, we started “depreciating” those emissions, so that the notional amount of emissions associated with the use of that transformer could be quantified and passed on to the purchaser of the copper that the transformer had helped produce.  This type of circularity is common in many value chains, and the E-liability method can capture it to make carbon-accounting calculations more accurate.

What’s the bottom line from the pilot on the choice of recycled versus virgin copper? The answer, of course, is: It depends. Recycled copper is not necessarily superior to mined copper in transformer manufacturing, in that the emissions associated with the processing of recycled waste can, as in this case, greatly exceed the emissions from mining copper, especially if the latter comes from sites that use low-emissions mining methods. This means that terms such as “circular” or “recycled” are not themselves indicators of emissions savings, and should be used with caution. The rigorous measurement of emissions specific to each input, process, and batch, at each step of a value chain, together with a rigorous accounting system for tracing those emissions as they move as inventories through the value chain, is the only reliable approach to determine which process is more carbon efficient.

Following from the pilot, Hitachi is now exploring ways to scale up the use of E-liability accounting across more suppliers and more products, to drive further decarbonization firmwide. This scaling up will mean subsequent iterations of the method can be still more accurate. It will, however, require technological solutions that dramatically lower the cost of implementation, which was largely manual during our pilot study. To that end, Hitachi has been engaging with a variety of software and data-assurance providers, such as SAP, that are developing E-liability inspired tools for large-scale adoption of these principles.