Julia Velkova is a post-doctoral researcher at the Centre for Consumer Society Research at the University of Helsinki. Her interests are in digital culture and media. She is currently involved in several projects on the politics and histories of emergent data infrastructures, with specific focus on data centres in the Nordic countries, and the labour, discard and temporalities that underpin their work. Her work has been published in journals such as New Media & Society; Culture Machine; Big Data & Society and International Journal of Cultural Studies, among others.

What is your current research about?

I am interested in the cultures and politics of media infrastructures, and in my current project I engage with the histories, temporalities and thermal politics of data centres that are being built in the Nordic countries. The topic that I have been working most recently on is the ways in which local municipalities and energy companies draw in the platform economy into energy politics, through the use of the thermal discard produced by data centres computing user data – a practice that exists currently in Sweden and Finland.

While we are still debating the implications of algorithms and data aggregation practices, there is a peripheral, and still quite marginal interest from scholars on the relation between digital media and energy. In this context, data centres have been the main target of criticism as they put pressure on local power grids while contributing with more carbon emissions, and at the same time reply on the use of water, which in certain cases is in very fragile ecologies, as Mel Hogan (2015) has written about in the context of Utah’s NSA data centre.

In the context of the Nordic countries, my interest has been not so much on the environmental implications, but on the ways in which data centres and the bodies, and living spaces of citizens living in cities are being reimagined by energy companies, city administrations and tech corporations as thermal infrastructure to cool down the commercial capturing and processing of data from Russian, Swedish or US internet consumption (see Velkova 2016 and Velkova, forthcoming). As the power position of energy monopolies is threatened by processes of decentralisation of energy provision in Europe, and by public concerns with climate change, new infrastructural and material arrangements need to be made to keep their power position. In this context, data centres, at least in the Nordic context, have surprisingly been reframed from pollutants to necessary future infrastructure to power cities with heat, a byproduct produced by consuming vast amounts of energy needed in order to compute data captured from everyday online practices.

I believe that in light of these infrastructural developments, we need a discussion about their regulatory, ethical and epistemological consequences in terms of our understanding of digital media, and particularly the materiality and politics of information and data.

How did you start to study the case of Yandex in Mäntsälä?

Somewhere around 2016 I read in the news that a data centre built by Yandex had become the main source of heat in a Finnish city – it sounded so exotic and extraordinary as an idea, that it captured my attention. At that time I was finishing my PhD in Sweden, and thought that researching more this practice would be a great future project. A couple of years later, I was lucky to get a position at the University of Helsinki as a post-doctoral researcher at the Centre for Consumer Society Research, and during my first year there I focused on trying to do empirical work in the Yandex data centre.

In comparison to many other companies which I had tried to approach earlier in relation to their data centres, Yandex were very open from the very beginning. I could visit their data centre in Mäntsälä – the city to which it provides heat – and interview the employees and the site manager at multiple occasions, including representatives of the local energy company – Nivos and the city administration of Mäntsälä. This research led me to the notion of thermopolitics of data which I am developing in an article manuscript at present. Thermopolitics is about the infrastructural mediation of heat through which bodies and spaces become simultaneously objects of quantification, commodification and of their own thermal regulation.

Based on my research, I suggest that thermopolitics of data is crucially contingent on practices of dematerialisation and disconnection, or the symbolic rendering of energy, bodies and spaces as immaterial, and unrelated to each other, a process which helps to install a new configuration of economic and social relations that can serve simultaneously computing machines, the platform economy and old energy monopolies, while not necessarily breaking apart from a carbon regime. In the formation of these relations, new regulatory, ethical and epistemological questions about the relationship between data, agency and energy arise, which also invite us to rethink our ideas about the relationship between information/data and energy.

How profitable is this heating project for Yandex?

According to Yandex it is very profitable, much more profitable than they had anticipated! They started this project as an experiment, but it turns out that it is one of the major sources of profit from this data centre. It is also economically profitable for the local energy company which has decreased its costs for heat provision, by reducing the use of Russian natural gas for heating and replacing it with the data centre generated heat. On the one hand, this seems like an economic and environmental solution which should be welcomed. At the same time, there are questions about ethics, regulations and the things which are not so vocally discussed.

For example, the focus on the use of a data centre as a heat source has brought attention away from the question of the source of its energy.  Less than ten percent of the electricity to power the Yandex data centre comes from Finnish renewable sources, while most of it arrives from the European grid as an energy mix of different fuels, which likely include coal and nuclear power. Hence, while heat produced by data computation might seem environmentally friendly in Finland, it actually leaves a carbon footprint somewhere else in Europe.

Another aspect is the ethics of these practices. What difference would it make if the citizens of Mäntsälä knew that they host, valorise and stabilise the exhaust of Russian data capturing and processing practices? Would it be morally or environmentally better to heat a city with the exhaust of US or Finnish data? Such ethical questions get suppressed by the silent integration of domestic space and bodily matter into the production of a data centre as a thermal infrastructure, and with it, the agency of the human bodies who are drawn in these arrangements gets also suppressed. While Yandex and Nivos have been very public about their practices, which are undeniably innovative, there has not really been any discussion or interest among the residents of Mäntsälä in relation to these practices, and the ethical questions that they raise. I find this surprising, because people care for where does there energy comes from, but they seem to stop caring when it comes from a data centre.

This in turn also raises regulatory questions – if Russian data heat can be a replacement for Russian gas, should its provision be regulated in some way too? The latter question is also related to the temporal dimensions of these heating infrastructures.

I have written already about the impermanence of data centres (see Velkova, 2019). Even though in the public imagination they seem to be durable and stable, there are multiple cases of data centres being shut down and opened in other places, which are though not often discussed in the media. In the case of Yandex, the currently shifting geopolitics of Russia in terms of its regulation and control over its domestic internet traffic also bring uncertainties to the future of the Yandex data centre in Finland, and what kind of services it could provide in terms of computing Russian online data. Hence, if data centres are shorter-lived than we think, arrangements such as those of Yandex might be also reshaping the temporality of energy infrastructures, shortening their life-span and making them disposable.

So, people in Mäntsälä are in fact dependent on Yandex.

Yes, at the moment they are dependent on Yandex for the heat provision in the city. Of course, the local energy company will find a solution if the data centre can no longer provide heat. Nevertheless, they do not consider it is very likely that an infrastructure such as a data centre can close in a moment. However, digital infrastructures are also shaped by shifting geopolitical arrangements, global politics and economy, which can make a data centre profitable in one location at a particular moment of time, and render it into waste in the other location or moment of time. In Helsinki, a data centre ran by the currently world largest provider of hosting infrastructure in data centres, Equinix, was shut down earlier this year despite contributing heat to the local district heating in the Finnish capital. It was shut down for economic and logistical reasons, only ten years after its inauguration. It suggests that data centres can be moved or re-allocated, and the heat infrastructure dismantled, if economy or geopolitics demand it.

How specific is the case of Yandex in comparison to other data centers? Is this exception or tendency?

Most companies that run data centres do not consider the heat that they produce as a valuable asset – they treat it as a waste. Yandex is in this context exceptional in terms of the scale in which it did valorise its waste heat in Finland. We hear often that data is the new oil. Yandex made sort of data exhaust a new gas. The use of heat from the Yandex data centre replaced the use of Russian natural gas that was used for heating in Mäntsälä. It is also exceptional in its scale – as it heats the town for most of the year.

However, as I said earlier, the use of heat raises other questions, and brings attention away from questions such as the energy sources, and the work of bodies and private spaces which are annexed to be part of the cooling infrastructure maintaining data computation.

In your recent talk at Aleksanteri Institute, you discussed data expansion of country to another country as new imperialistic order. 

In this talk I was mentioning some of the scholarly debates on datafication and data infrastructures, which have related data and media infrastructures to questions of empire (Aouragh and Chakravartty, 2016; Rossiter, 2017) which are going on in parallel to the debates about the environmental impact of digital media.

The point that I wanted to make in this presentation was that we usually consider geopolitics as a state-driven process. However, infrastructures also install forms of governmentality, outside of the traditional political institutions (see Larkin, 2013). Data centres such as Yandex also operate from and manage land and physical territory, but they also connect and disconnect populations, change configuration of everyday practices and economies. As Keller Easterling (2016) writes, infrastructural space can operate as a form of extra statecraft. When a data centre like Yandex operates with energy, land and data about people, it instantiates forms of power, territoriality and politics independent of the state, yet shaped by state policies, of course.

Are data centers and data infrastructure actors getting more powerful in algorithmic reality?

For me the question of power is that their power is expanding in new domains, creating new forms of convergence. Companies such as Yandex or Google have dominant positions in the regional or global markets. But they do not only operate with data, they are also reconfiguring urban infrastructures, entering into energy politics, transport, and multiple other domains of everyday social life. In this process, as José van Dijck et al (2018) convincingly wrote in their book on the platform society, what is at stake is public values, and of course, the formation of ever stronger monopolies who manage not just our data, but who mediate our everyday life. While we take pleasure and convenience of new technologies and infrastructures, our mediated everyday life increasingly brings profits to a handful of companies through each activity we do, in ever expanding domains of life, getting now to the critical urban infrastructures of energy provision.

I think also that the Yandex data centre in Mäntsälä allows us to reconsider the ways in which digital media, information and the so-called, ‘knowledge economies’ can be considered post-industrial. We have grown used to the idea that information, and today – data – is immaterial, running only through signals without a place and physical infrastructure in the world. But as the technologies which need to maintain data capturing and computation practices get bigger and ever more connected to questions of energy, environment and society, we must rethink and reconsider the extent to which we have not become post-industrial, but remained heavily industrial.

References:

Aouragh M and Chakravartty P (2016) Infrastructures of empire: towards a critical geopolitics of media and information studies. Media, Culture & Society. DOI: 10.1177/0163443716643007.

Dijck J van, Poell T and Waal M de (2018) The Platform Society. New York: Oxford University Press.

Easterling K (2016) Extrastatecraft: The Power of Infrastructure Space. Paperback edition. London New York: Verso.

Hogan M (2015) Data flows and water woes: The Utah Data Center. Big Data & Society 2(2). DOI: 10.1177/2053951715592429.

Larkin B (2013) The Politics and Poetics of Infrastructure. Annual Review of Anthropology 42(1): 327–343. DOI: 10.1146/annurev-anthro-092412-155522.

Rossiter N (2017) Imperial Infrastructures and Asia beyond Asia: Data Centres, State Formation and the Territoriality of Logistical Media. The FibreCulture Journal (FCJ-220).

Velkova J (forthcoming, under review) Thermopolitics of Data: Cloud Infrastructures and Energy Futures. Cultural Studies.

Velkova J (2016) Data that warms: Waste heat, infrastructural convergence and the computation traffic commodity. Big Data & Society 3(2): 1–10. DOI: http://journals.sagepub.com/doi/full/10.1177/2053951716684144.

Velkova J (2019) Data Centres as Impermanent Infrastructures. Culture Machine (18). Available at: www.culturemachine.net.

 

Leave a comment

Your email address will not be published. Required fields are marked *