Sustainability in the semiconductor industry

Why evaluating the impact of electronic based systems (EBS)

For a long time the electronic industry had a reputation of a clean industry, partly because of its manufacturing environment (cleanroom, highly pure raw material), partly because the tiny size of EBS products makes their impact appear negligible, compared to that of bigger parts. However, in recent years the direct environmental impact of this industry has been investigated more closely, with a particular focus on four aspects:

  1. The source of raw materials, many of which are critical, that is they are scarce and at the same time economically and technically important, or are extracted in terrible social conditions
  2. The massive use of chemicals, many of which are toxic or with high GWP factor
  3. The high energy consumption, both during production (due to cleanroom environment, high purity water and chemicals, high temperature treatments) and during use. Today, digital services alone are responsible today for 4% of the overall CO2 emissions, and their carbon footprint increases 9 % yearly)
  4. the largely unsolved issue of recycling and disposal of electronic products.

The direct effects are the easiest to assess, they, so to say, hard facts. But in the case of EBS, the impact extends well beyond the product itself. Indirect effects are linked to the de-materialization effect of ICT and to the improved efficiency of production and logistic. For sensing technologies, energy management, IoT those aspects are crucial.

But EBS in digital products heavily influence the habits of the users and the society itself through e-commerce, social media, smartphones and other mobile devices. This leads to the so-called structural and behavioural effects of ICTs. A typical example is the rebound effect, which occurs when the resources released by efficiency gains are not saved, but employed for some other activities (for example, the shrinking size of electronics led to the integration of more functions in smartphones, not to smaller, environmental-friendly phones).


The challenges of modelling the sustainability of EBS

The importance of indirect and structural effects is certainly one of the specialties when modelling the impact of EBS. Although it is not mandatory to include those effects in the analysis, they certainly unleash the biggest potential for innovation.

Besides the methodological challenges, some practical matters make the adoption of life-cycle thinking in the semiconductor industry difficult. The data availability is made worse by two characteristics of the this sector: the short technology life cycles and the global, interlinked supply chain. On the other side specifics of this industry could support LCA adoption in this sector:

  1. processes and materials, often based on jointly developed technology platforms, are similar also across manufacturers and technology generations, literature data from existing studies can be used for an estimation of the inventory in other technologies, and compensate in this way to the lack of data;
  1. Manufacturing processes are built up out of modules, that, just like building blocks, are employed again and again in different processes. Once the single modules are modelled, these data can be re-used in many impact analysis studies;
  2. the supply chain is fragmented, but also interlinked, and many companies are used to cooperating, in order to share the huge development investments and risks. This attitude could be used to deal with the challenges. While single enterprises cannot exploit the positive aspects, industry associations and joint initiatives could develop the methodology and suitable databases to support the diffusion of life cycle approach in the industry.


How to measure the sustainability of EBS

What is applicable for EBS, and what do the peculiarities mean for measuring the impact?

LCAs are an excellent method to calculate the environmental impacts, especially if you limit the analysis to the carbon footprint of your products. Software tools and databases are already well developed (although not specific enough for an accurate modelling of EBS), and can be customized to allow systematic studies of product portfolios, taking advantage of the modularity of manufacturing processes in electronics. Even a linking to the established EPR system for "real-time LCAs" is thinkable. You can expand the system to consider the end products, gaining insights for innovation and marketing. But they are limited to environmental aspects and, more critical for EBS, indirect and structural impacts are more difficult to include.

Better suited for that, as holistic in their approach, are the SDGs. Choosing the appropriate indicators you can measure and manage the impacts of your activities in a comprehensive way, including the indirect and the structural aspects. But they are too generic to allow comparisons between products or technical solutions. For example, if your company produces energy management systems with less heave metals than your competitors, you can measure your impact on SDG 7 (affordable and clean energy), SDG 14 (life below water), SDG 15 (life on land) and possible other SDGs, but you cannot directly compare your solution with other ones. Nevertheless, SDGs are an excellent tool to manage sustainability as a whole and are increasingly demanded by stakeholders of all kinds.

And what about circular economy? Recycling and resource scarcity are taken into account in LCAs, but their contribution is melted in the pot with all other aspects. So, if your company is actively recycling production rejects, you might want to highlight that with additional indicators such as the Material Circularity Index, ideally to integrate the LCA analysis (since they are calculated from similar sets of data). The fragmented supply chain and the position of EBS in the middle value chain restrict the availability of data and the freedom to operate to improve those indexes, but, given the growing political pressure and the scarcity of resources, it might be a worthwhile addition to other methods.