Reproblematising maintenance,reconsidering infrastructures: The emergence of a more-than-corrective maintenance programme in French water networks management

Questioning infrastructure maintenance

Academics have long been more concentrated on infrastructure extension and its future-oriented dynamics than on the management of existing infrastructures (Florentin, 2018; Londong, 2003; Scherrer, 1992). The emergence of a certain vision of modernity, largely resting on the ideas of rational, comprehensive, unitarist and standardised planning and networked technologies to tackle uneven development (Mattelart, 1994), has been concomitant with the construction of urban technical networks and of intensive waves of urbanisation (Coutard, 2010). This urban translation of the belief in the “technological sublime” (Nye, 1994) forms what has been coined “a modern infrastructural ideal” (Graham and Marvin, 2001).

Such an ideal has been at the centre of numerous critiques. A prominent one claims that the extension model is problematic because it has been led astray. Important works have underlined how the degradation of welfare state combined with other spatially selective processes have made the extension of numerous infrastructures synonymous with aggravated socio-spatial inequalities and contributed to the capture of common goods and natural resources (Loftus, 2006; Swyngedouw, 2015). A similar “developmental” perspective is at play in research that has focused on infrastructure deployment and extension programs in the global South. Here, the critics share a Foucauldian perspective, as they highlight the problematic consequences of the promises conveyed by infrastructure, especially in terms of governmentality (Hetherington, 2014) and the reconfiguration of subjectivities (Baptista, 2018; Pieck, 2013). Somehow following the work by Akrich (1992) and Mukerji (2009), this scholarship describes infrastructures as one of the fundamental means for the State to ensure its sociomaterial presence (Harvey and Knox, 2016). Infrastructures are here apprehended as political agents that have the ability to (re)order spatial, economic, social and even moral relationships (Fairhead, 1992; Summerhill, 2006). Under the guise of rationalisation and universality, two pillars of modernity, they generate inequalities in regions they dramatically reorganise (Alvey, 2014; Hetherington and Campbell, 2014).

For more than a decade now, a growing body of research work has been inviting to decentre such an approach, to explore dimensions of infrastructures that are overlooked, when not invisibilised, by the focus on extension dynamics. This displacement notably occurred through the investigation of “disrupted cities” (Graham, 2010), and the reminder that infrastructures experience a constant risk of failure (Amin 2016). In this Heideggerian perspective, instead of being questioned only as a network whose growth and stability participate in ordering rural and urban life, infrastructures are apprehended as a complex assemblage whose heterogeneity and vulnerability are exposed through the lens of breakdown (Bennett, 2010). Vulnerability is what defines modern infrastructures from this standpoint, and a crucial dimension of their contemporary governance.

The decentring is even more marked among the burgeoning “maintenance and repair studies” (Denis and Pontille, 2020) In their seminal paper, Graham and Thrift (2007) insist on the importance of the manifold interstitial, and mostly mundane, interventions that, far from the abstract idea of seamless extension, “actually allow [infrastructure systems] to sustain the distanciated connections and flows that they are designed to deliver” (p. 10). Maintenance and repair studies do not solely focus on the development and inertia of infrastructures, but rather on their “life”, a life that is not as fixed and monotonous as the developmental perspective may assume. Infrastructures are maintained and repaired in a continuous manner (Denis and Pontille, 2015; Henke and Sims, 2020; Jackson et al., 2024; Vinsel and Russel, 2020), far beyond the temporal scope of exceptional breakdowns (Denis and Pontille 2025; Graham, 2010). Such an approach, which questions how infrastructures constantly and continuously “hold” and endure (Denis and Florentin, 2022b), consists of keeping an eye at infrastructures after they have been deployed and extended, and taking their longevity as an object of research in itself, rather than a given. In other terms, it favours a temporal lens over a spatial one (Barry, 2020). But instead of insisting on crisis and disruptions, it leads to highlight the generative force of continuity, its political implications, and explore infrastructures as things that are never completely stable nor finished, as sociomaterial assemblages that constantly change to remain the same (Denis and Pontille 2025), and whose becoming is tightly intertwined with the one of the people who take care of them (Cohn, 2017).

These dimensions have long been at the centre of a highly influential scholarship that explored the modes of existence of infrastructures in the global South, especially highlighting their heterogeneity and precariousness (Jaglin, 2014; Lawhon et al., 2018). These works demonstrate that infrastructures “hold” despite their fragmented and sometimes extremely unstable condition thanks to a great variety of interventions. Many of these interventions are mundane, and carried out by diverse people, often largely beyond the scope of engineers and specialised technicians (Anand, 2017; Simone, 2004; Wahby, 2021). Importantly enough, they show that infrastructures in the global South are never completely installed, or “completed”. They are rather always “in the making” (Baptista, 2018). This means that they need to be questioned—as well as the sociomaterial order they participate to shape—by breaking away from the “completist lure” (Guma 2020) which still drives a lot of infrastructure studies. This also calls for turning the logic of the neo-foucaldian critique on its head by providing a more realistic depiction of the processes that drive “modernity”: far from being intrinsically powerful and doing “politics by other means” through their material sturdiness and obduracy, infrastructures prove to be fragile (Velho and Ureta, 2019). Taking this fragility seriously, and thinking “with” it, offers promising avenues to a renewed critique (Domínguez Rubio et al., 2025). For instance, the consequences of infrastructure fragility in terms of social and spatial inequalities cannot be reduced to a matter of accessibility and non-accessibility. They rather unfold on a continuum between those who can benefit from consistent and reliable infrastructures and those who have to deal with (and generally take care of) the most deteriorated, or “uncompleted” ones. This notably raises crucial questions on labour issues (who concretely works to keep infrastructures functioning? At what financial and physical costs?), and gender inequities (Alda-Vidal et al., 2023, Ramakrishnan et al., 2021a).

Yet, reducing the conceptualisation of infrastructure fragility to studies in southern configurations would be fundamentally misleading. One of the main stakes associated with the emergence of maintenance and repair studies lies in de-essentialising some sociomaterial properties of infrastructure in the global South, by foregrounding the specific fragilities of northern networks, especially sensible over the last decades:

With the liberalization, privatization and ageing of infrastructure in the global North sometimes leading to substantial interruptions of service, coupled with a widespread withdrawal of profit-draining back-up systems, the always oversimplified distinctions between the cities of the global North and South are beginning to lessen. (Graham and Thrift, 2007: p. 11)

Here again, it is thus a matter of continuum rather than an intrinsic distinction. Heterogeneity, precariousness, and incessant mundane interventions are important features in global North as well, even though, of course, they do not unfold at the same scale and have different social and political implications.

From South to North, most of the works that insist on taking maintenance and repair into consideration share another important aspect. They are attached to the situatedness of the configurations they investigate and never wander away from the practicalities of infrastructures management. Favouring case studies, they avoid questioning infrastructures in abstract terms and generalising their results in overly schematic fashion. This allows both to foreground the rich relationships between maintenance and repair and the situated ethics of care (Puig De la Bellacasa, 2017), and to acknowledge the irreducible variety of configurations that involve infrastructures maintenance, notably in terms of temporalities (Ramakrishnan et al., 2021b; Röhl, 2023) and modes of knowing (Denis et al., 2023; Solé-Pomies, 2023). Such attachment to local cases is also a way to highlight the ambiguities and ambivalences of infrastructure maintenance. These studies take into consideration the struggles that can unfold around situations of disrepair (Chu, 2014), as well as the power dynamics at play in the distribution of maintenance work itself (Barnes, 2017). Through the analysis of contrasted and empirically rich cases, most maintenance and repair scholars strive to avoid the pitfall of romanticism and refuse to flatten disparities and frictions through disincarnated uses of notions of repair (Ramakrishnan et al., 2021b) and care (Dominguez-Guzman et al., 2022; Duclos and Sanchez Criado, 2019).

Finally, and consequently, this literature invites us to break away from any kind of evolutionist vision which would depict maintenance as a uniform phase in the development of infrastructures, a phase naturally associated with an “ageing” of infrastructures that could be defined thanks to universal and stable criteria. Maintenance (and the lack thereof) is a more or less publicly discussed matter of concern. It becomes questioned in specific configurations, which crystallise particular moments in the lives of infrastructures (Denis and Florentin 2024).

We are currently witnessing one such moment in numerous global North countries, where recent dramatic breakdowns have been debated in terms of both aggravated decaying of urban infrastructure and deferred maintenance (Henke and Sims, 2020; Vinsel and Russel, 2020). In various cities of the global North, infrastructures seem today less and less discussed with regards to their sufficient or insufficient extension as to their decay, the hidden or underrated costs of their upkeep, and the urgency of their repair ¹ . In this paper, we propose to explore some of the practical transformations entailed by this reproblematisation, which we consider a crucial, yet overlooked, aspect of the contemporary “age of infrastructure maintenance” that numerous “rich” countries are experiencing. First, we show that the reproblematisation at play in Europe and France frames maintenance differently, spanning from a set of reactive repair interventions to what we call a more-than-corrective framework. In such a framework, preventive and anticipatory practices are added to the routine reparative operations. Then we focus on three transformations generated by this widening: the reorganisation of infrastructures’ description; the documentation and supervision of infrastructures’ life; and the renewal of infrastructures’ valuation practices. These three transformations contribute to both a reordering and a reconsideration of infrastructures.

To investigate such reconfigurations, we will focus on French water networks, the management of which has recently been the object of important reorganisations, which epitomise this particular moment. Through this inquiry, we want to highlight a series of practical, political and ontological implications of the emergence of a more-than-corrective infrastructure maintenance in a particular setting. This is to be considered a first step into questioning the new role maintenance is (or is not) progressively taking in infrastructures management in recent years.

Case and methodology: Ethnographic foray into the French water sector reconfigurations

For several years now, French water services have been experiencing important regulatory, institutional and organisational transformations, which have profoundly affected the way infrastructures are managed and handled by their different stakeholders. This transformation can be summarised as a twofold one.

First, a dramatic restructuring of the local governance of infrastructures has taken place under the aegis of a new law redefining administrative boundaries for various public services (loi Notre, adopted in 2014). Until then, water management was carried out at a variety of territorial scales, including very small municipalities. This is gradually shifting to a larger and uniformised level (intercommunal) where the communities in charge of the networks are recomposed, partly to get larger engineering competences and investment capacities.

The second ongoing transformation has to do with time and environmental regulations. At the national level, there is a growing concern amongst both public and private stakeholders regarding the age and state of infrastructures, as an important part of them was installed 70 to 80 years ago if not more. They are considered a potential threat for the continuity of this urban service and their decay is seen as a risk for the environment. This concern, as we will show further, has led to the implementation of new requirements in terms of supervision and upkeep practices, directly linking environmental issues to maintenance activities.

From a methodological standpoint, these transformations offer an opportunity to investigate water infrastructures management practices while they are being reconfigured, not yet stabilised, even less routinised. They thus represent a unique occasion to understand how the problematisation of infrastructure through maintenance is being simultaneously experienced and enacted by numerous stakeholders.

What does it take concretely to put maintenance to the fore? What operational and organisational challenges does it raise? How are infrastructures themselves reconsidered in this process? To answer these questions, we conducted an 18-month in-depth study of water services in France, nourished by diverse empirical material, including 102 semi-structured interviews with different stakeholders in diverse territorial configurations (metropolitan, semi-rural, mountainous, coastal, formerly industrialised, etc.), observations within companies, on sites of operations, in several meetings of the 2019 National Conference on Water (Assises de l’eau) and in two national trade fairs dedicated to water management. Our approach is inspired by the contributions of institutional ethnography as they have been detailed and discussed in Dorothy Smith’s works (1987, 2005). Such an approach is built on an in-depth understanding of work processes and practices, which are established as a starting point for investigation and analysis. This has three main consequences. First, to adopt such a perspective is to consider that the persons studied are experts on the situations in which they are immersed, which means their vocabulary, the way they characterise the problems they face, and the solutions they collectively and locally assemble are to be taken seriously. In other terms, contrary to a structuralist or institutionalist perspective, we cannot start from pre-stabilised categories, may they come from an academic, legal, regulatory or managerial background, to account for the practices we study. It is precisely the way people struggle with these categories, arrange them, discuss them, or create new ones that is placed at the centre of the investigation. From this standpoint, management and governance are studied in their practicality and mundanity (Woolgar and Neyland, 2013).

Second, rather than simply putting these legal and regulatory frameworks aside, institutional ethnography invites to meticulously explore the ways through which rules, standards and norms “intervene” in specific situations, and how they are concretely dealt with, notably through the production, circulation and manipulation of various textual artifacts. Instead of seeing rules as being “applied” or not, or procedures as “followed” or not, this approach strives to pay attention to the way “regulatory frames” (Smith, 2005) are processually enacted in very concrete things such as mundane conversations, official documents, technical texts or organisational processes. In our case, this means that, in lieu of invoking the general concern for infrastructure maintenance and the extended maintenance programme that was emerging through legal texts and official guides as “contextual” elements that can explain the nature of the practices we were documenting, we paid attention to the concrete problems and specific challenges such changes would raise locally. We investigated how national documents, local ones, and mundane practices, altogether, were shaping up a new place for maintenance in water infrastructure management.

Finally, and consequently, conducting an institutional ethnography amounts to examining the concrete conditions for the coordination and cooperation between stakeholders, without making it an analytical prerequisite for collective work. Questioning the ways work is distributed, organised and re-organised is crucial to identify and understand the sometimes subtle operational and conceptual shifts and micropolitics that occurred within water management and the possible emergence of new facets and figures of infrastructure management.

The emergence of a more-than-corrective maintenance programme

In Europe and in France, series of documents, training schemes and policy programmes have designed over the last decade something that could be considered a new doctrine. This doctrine puts an extensive form of maintenance at the core of strategies for sustainable infrastructure management. The French has its own term to encapsulate most of the concerns and issues raised by infrastructure maintenance, “gestion patrimoniale”. Even though it may be translated into “asset management”, the notion, as we will further explain, carries a more subtle meaning, which goes beyond financial and accounting matters, and addresses issues of national and industrial heritage. Importantly enough, the expression conveys a broad definition of maintenance, whose ambition aims at expanding infrastructure management well beyond the accumulation of corrective operations in reaction to more or less dramatic breakdowns.

As the regulatory texts that now form the backbone of French environmental laws were published ² , what was intended by gestion patrimoniale of water infrastructures became more explicit. These texts were completed by workshops and guidebooks dedicated to water infrastructures maintenance. At least two main elements can be identified from this material that help characterise the type of maintenance that is entailed and how it has been conceptualised: the production of a doctrine and its declension through specific operations.

First, the maintenance of water networks became, throughout the publication of these official texts, a large-scale problem that needed to be addressed through new national measures. In these texts, maintenance, under the umbrella of gestion patrimoniale, appears as a specific concern, and is problematised at the crossroad of health and environmental issues. This concretely translates into a series of key operations that have been designated as mandatory for every local authority, with a penalty system that has been established in the event of non-compliance (doubling the fee local authorities owe to the “catchment basin agency”—agence de l’eau—under which they operate). At the same time, these operations have been turned into performance indicators, displaying, amongst many items in a dedicated database (SISPEA), the progress of gestion patrimoniale at the national level.

Amongst these compulsory operations, two appear to be the cornerstones of the doctrine which has gradually been stabilised in the form of a more-than-corrective maintenance framework.

The first one is the production of a detailed and updated description of the water network and all its components. This may appear as a very simple and modest operation, at least if one imagines that every authority already has access to a set of exhaustive documents, maps and data that provides such a description. And indeed, this assumption is partially shared in the regulatory texts themselves, which more or less explicitly present such descriptive operation as a precondition to the set-up of a more-than-corrective maintenance programme. Yet, a detailed and updated description of the network is anything but an easy thing to obtain. And the fact that this is pointed as a crucial, even though preliminary, component of gestion patrimoniale is particularly emblematic of the reproblematisation at play. Far from the simple accumulation of reactive technical interventions, infrastructure maintenance now ought to be considered as a multifaceted domain, in which building a comprehensive knowledge of the network, apprehended as a whole, takes an important place.

The second pillar of the doctrine points towards the temporal organisation of maintenance practices and their scope. Gestion patrimoniale, as the guideline published by ASTEE in 2012 states, “is a long-term approach that takes into account the state of the asset (patrimoine) all throughout its life cycle” (p. 11). This has at least two important implications on a practical level. First, such a long-term approach implies that the pace of interventions is rethought, and at least partially detached from a strictly reactive logic of repair. The very idea of more-than-corrective maintenance lies in the ability of infrastructures managers to go beyond the mere handling of daily crises and breakdowns. This practically means that infrastructures managers need to perform a constant supervision of every part of the network. Such redeployment of some of the work activities allows to establish a sociotechnical routine that enables the daily cultivation a new kind of attention towards infrastructure (Denis and Pontille, 2023). Second, more-than-corrective maintenance is enacted through the establishment of pluriannual investment programmes. Gestion patrimoniale is a matter of planification, and infrastructure managers now have to organise maintenance work within a wider timeframe. In addition to investing into day-to-day monitoring, they use a variety of calculative devices that should allow them to discuss, evaluate and prioritise maintenance interventions, to plan their scheduling with the best possible control of costs, conditions of works realisation, and risks of breakage.

This temporal aspect of gestion patrimoniale is tightly connected to the previous one, obviously. Knowledge, routine supervision and anticipation go hand in hand, and feed each other.

The framing of these documents does not mean that new management practices will be mechanically put in place, though. However precise and restricting they may be, official texts never constitute sufficient resources for action. They offer landmarks and references, but are also sources of practical difficulties. The hardest part is still ahead, so to speak: managers and technicians have to enact gestion patrimoniale. What does it take exactly? How the principles and the compulsory operations that are identified in the doctrine are enacted?

In the following sections, we show that the two aspects we described generate operational, organisational and sociotechnical challenges within the utilities, which leads to a profound reconsideration of infrastructures themselves, as they became subject to a new scrutiny in the perspective of a more-than-corrective maintenance. First, the need to produce detailed descriptions of the network translated into situated enquiries during which whole parts of infrastructures were (re)discovered. Second, the establishment of a new perimeter of maintenance went hand in hand with the reconfiguration of knowledge production, notably including a “biographical” dimension. Third, and lastly, this reconsideration leads to questioning the different values that are embedded in infrastructure and its components and engage managers and technicians in renewed valuation practices.

(Re)discovering infrastructures, reorganising services

The idea that utilities must have a detailed description of the network they oversee is at the heart of all recommendations. It seems a reasonable and fairly obvious matter to condition interventions, their prioritisation or the long-term maintenance planning beyond reactive repair operations, to have an accurate and comprehensive view of the network. Yet, such overview is anything but easy to produce, especially when it comes to underground networks. And, even though inventories were depicted as a prerequisite for the deployment of a fully-fledged gestion patrimoniale, most managers were actually lacking such detailed view.

Indeed, the first attempts at producing inventories in French utilities highlighted how partial and fragmented the knowledge of networks actually was. And, most importantly, they made everyone realise that filling the gaps represented a far more sophisticated operation than simply mechanically circulating information from local to central departments. In numerous places, information was simply missing. Before being able to evaluate the state of the infrastructure they managed and make plans for future works, managers faced very simple questions, which they actually could not answer: “What are the pipes and installations made of?” and even “Where are they exactly?” or “How many structures do we have?”

[We have] a few boroughs in the mountains. They have sometimes three, four sources per borough. So, a very disparate situation between the areas. And a total ignorance of all this. (...) We had pipes passing in the mountain, in the countryside, sometimes underground, but sometimes directly on the ground, in a state I cannot even describe… They were dating back from the 19th century, sometimes in a non-protected material… In many occasions, we did not even know where they were circulating. (Interview with the head of gestion patrimoniale department, December, 2018)

From a national perspective, in an era of Big Data, Artificial Intelligence and the multiplication of sensors, addressing such issues seems to be a very simple thing to do. But at the local level, getting to know the infrastructure has been and still is a real challenge.

This challenge concretely and organisation-wise translates into the creation or reinforcement of departments specifically dedicated to this knowledge production, or to the transformation of the current organisation to give a larger room to all the activities entailed by the implementation of more-than-corrective maintenance. The problematisation of infrastructure through maintenance not only transforms the types of activities, but also the organisation of the various utilities, becoming in some cases their central strategy. In the French context, the emergence of these (new) jobs and activities within utilities or local authorities is also accompanying the development of new competences within local authorities around infrastructure management, whereas the traditional engineering capacity developed by the national State is progressively disappearing (Barone et al., 2016). In that sense, the rediscovery of infrastructure through maintenance is also a time and a window of opportunity for re-regulating water locally.

In the metropolitan area of Nice, like in all the cases we have explored, managers and employees of the local utility have faced enormous challenges to reconstitute the past of the water network over the last 5 to 10 years. This is all the more difficult in a supplied area that is now twice as big as it used to be, due to the territorial transformation of water management accelerated by the loi Notre which makes water an intercommunal matter.

Far from a mere technical prerequisite, thus, the very description of infrastructure engaged managers in the production of maps and surveys, which proved to be operations far more time-consuming than initially envisaged. In Grenoble for instance, the production of an inventory took 5 years, while in the Nice area, the manager evokes a real “jungle” when explaining to us what his teams had to accomplish. He is furthermore unable to calculate the cost of the whole operation, which he was still considering unfinished in 2018, several years after it started. Similarly, in Lorient, the first part of the inventory took 5 years and still requires constant actualisation.

Rather than externalities or poorly anticipated collateral issues, these infrastructural enquiries can be seen as the first significant operations of a broadened maintenance programme, no longer limited to corrective measures. Indeed, far from simply “filling the gap” in terms of information, they lead to a real (re)discovery of networks within utilities. Such rediscovery owes much to the specific movement the enquiries allowed: they constituted opportunities for managers and technicians to initiate or renew a form a proximity with the infrastructure they were in charge of. By getting into direct contact with the network materials, they were able not only to clarify existing information, but also to spot components that were forgotten, or simply unknown. For, just as buildings (Yaneva, 2001), infrastructures can “surprise” during such situated encounters. While we accompanied a maintenance technician in the south of France during one of his on-site visits, we witnessed such (re)discovery process when we saw him counting four structures around a reservoir instead of the three that were represented in the official map he was holding in his hand. This was not an unusual finding, he told us. Since the beginning of his regional inventory assignment, not only had he corrected a great deal of information about the location of pipes and structures, but he also had regularly discovered elements that were nowhere to be found in the available information systems (on paper or digital), even though the information system of the utility is known to be amongst the most detailed in the French context.

In that sense, the inventories that have been conducted to initiate a more-than-corrective maintenance programme must be understood as generative. They install the production of knowledge not so much as a precondition for “good” maintenance, but as a dimension of maintenance itself, which dramatically extends and enriches the way infrastructures have been apprehended within utilities. This reconfiguration has also a political dimension, since it displaces the role and agency of public actors dedicated to infrastructure management. Such programme especially reinforced the expertise of the utilities (and of the local authorities owning the infrastructures), where a new kind of engineering has progressively taken shape while ever more detailed knowledge of infrastructures was gained. This, as we will see in the next section, is particularly striking when considering the kind of information gathered during these various enquiries, beyond the questions of composition and location only.

Documenting the life of infrastructures

Knowledge and maintenance are articulated in many ways. Obviously, the modes of knowing (Law, Ruppert, 2016) deployed by maintainers dramatically change depending on the setting. Knowledge required to keep a bridge in a functional shape largely differs from the interventions to maintain electric grids. Modes of knowing may also vary within a same domain (Solé-Pomies, 2023). Furthermore, the very object of knowledge production also differs in maintenance practices. As Denis et al. (2023) have shown, maintainers do not always consider the same dimension of a thing when they try to take care of it. Depending on the situation, these practices may strive to document and understand at least three different things: functionalities, behaviours, or biographies. In the case of urban infrastructure such as French water networks, functional knowledge—which mainly comes from engineering—and behavioural knowledge—which is provided by material sciences and chemistry—are relatively stabilised and have not been directly affected by the turn towards a more-than-corrective maintenance. On the other hand, in numerous utilities we studied, the situated enquiries and the renewed proximity they allowed were invested as ways to gather hitherto neglected biographical information on infrastructures. Getting closer to the pipes and the installations helped to know more about the events that punctuated their life: dysfunctions, breakdowns, repairs, interventions of all sorts….

To do so, the direct observations were completed by the gathering of all the archived documents that could be found, from paper maps to official reports, contracts or technical specifications. The on-site visits were also occasions for maintainers from the central services to meet with local maintainers who had been taking care of parts of the network for years and detained more informal information.

During the visit we shadowed, after he inspected the site—locating the structures and reckoning their state, observing, touching and measuring different elements—Marc, the technician, turned towards the two people who had emptied the reservoir and cleaned it. They had been maintaining the infrastructure for a long time, having worked for the municipality before being employed by the public utility. They knew a lot, and were keen to answer their visitor's questions, recalling former breakdowns and interventions, transformations and small adjustments in the form of stories depicting the various events that have rhythmed the reservoir’s biography. Earlier on, Marc had stumbled upon printed maps in one of the structures, and had taken pictures of them, so he can add them to his files (Figure 1).

OPEN IN VIEWER

Figure 1.

Photographing a map to consolidate infrastructure knowledge (photo: J. Denis).

A few days later, he recorded the information he gathered while visiting the structures, as well as the pictures and the notes he took during the interviews in his database. He translated them into proper alphanumerical data (e.g. dates and types of intervention) or kept them as freeform comments. They would complete what he had already gathered in the archives and the previous information system. The current database, which he custom-designed together with his hierarchy, is one of the main instruments of the gestion patrimoniale in the utility, helping everyone to identify urgent repairs to be done, anticipate further less urgent interventions and build a pluri-annual investment plan.

Such knowledge production consists, thus, in apprehending the infrastructures as living entities, which experience different kinds of events that are important to know to feed maintenance processes. This is thanks to such biographical knowledge, turned towards the past, that maintenance can escape the fate of the repetition of corrective reactions and be deployed towards the future, through anticipations and planning schemes.

That said, getting a detailed description and diagnosis of networks is not just a past-oriented practice that will produce a fixed and stable picture of the different components of an infrastructure once and for all. Otherwise, a similar issue of partial ignorance would emerge regularly, leading to new time-demanding and costly campaigns of enquiries to produce an updated version of the network. Even biographical knowledge needs to be kept up-to-date. Another tool for a more-than-corrective maintenance thus lies on the organisation of information flows that allow such updates. In addition to inventory campaigns, some utilities have passed the requirements of the network description directly onto the technicians that are in daily contact with the network and its not-so-peaceful life: maintainers themselves are instantiating maintenance interventions as opportunities to generate data. Identifying an error of geolocation after uncovering a portion of the network for a repair or replacement operation, finding the yet unknown material composition of pipes, and of course recording the date and nature of actual interventions, constitute invaluable information which could not nurture collective knowledge about the infrastructure, would these elements remain stuck at the scale of the local intervention. As a technical director working in the vicinity of the Annecy Lake explained to us:

This is an everyday challenge, you know. First and foremost, we need human presence on the field and feedback from it, and with tablets on the field if possible. A layer of information, of observation, which is being deployed. There, we will have geometric corrections, saying for instance: “this pipe is not well positioned on the map). We make some detection, and bingo, we have it readjusted.” This is one layer of observation. We do the same with leakage or with hydrants, to position them and fix mapping errors. (Technical director of a water utility in the Alpes)

This is a shared concern among utilities which has been at the centre of many innovations in recent years. Several companies offer technical solutions to facilitate the process of data reporting coming directly from the field of operations.

The advent of wireless technologies and handheld devices such as smartphones and tablets has radically transformed the possibilities for field workers to generate information about their own work and integrate it seamlessly into the information systems of the organisation for which they work. In the case of water utilities, the organisational transformations entailed are far from negligible. First of all, they consisted in an unprecedented way of involving maintenance workers, who were asked to develop new attentional skills. This can be seen as a recognition of the situated knowledge of workers who are generally undervalued, and who are now considered to be the most appropriate people to contribute to the knowledge of the network. But this should not be considered a benefit for everyone. To some maintainers, this may represent a new burden, especially for those who are uncomfortable with digital technologies and data handling.

Practices entailed by these more-than-corrective maintenance programmes highlight a specific relationship to knowledge production, which fully takes on its intrinsic incompleteness. They show how maintenance and infrastructures are problematised and politicised anew through knowledge production issues, far from a mere acknowledgement of infrastructural decay. The investment in information gathering and data generation does not altogether aim at exhaustivity. Even more, it does not assume that exhaustivity could occur in the first place. Acknowledging the obvious inaccessibility of a mostly underground infrastructure and the inevitability of its incessant transformations, more-than-corrective maintenance programmes draw on the situated encounters with the components of the networks, which technicians experience repeatedly, to cultivate an open enquiry-based approach and articulate it with increasingly evolutionary systems of geographical information. Such an approach embraces a general principle of uncertainty, and symmetrically, recognises the necessity of never stopping the investigations. As part of such maintenance programme, knowledge is no longer conceived of as a mere stock that needs to be rebuilt or restored, but the evolving result of an ongoing daily technical and organisational process. In this perspective, knowing is a matter of continuity and thus constitutes a fully-fledge component of maintenance. The production of knowledge is therefore at the core of the transformations of infrastructure management at play.

Re-valuing the infrastructure

The reconsideration of infrastructures that goes alongside the emergence of a more-than-corrective maintenance also encompasses new or renewed valuation practices.

Some scholarship in infrastructure studies have argued that the value of infrastructures was less and less residing in the materials and increasingly in the flux (Rutherford and Coutard, 2016). This trend seems to be disturbed by the actual reproblematisation of infrastructure maintenance, in which utilities are rethinking their infrastructures not only as an abstract asset, but also a legacy and a form of heritage that needs to be taken care of. Two facets of this transformation can illustrate how infrastructure is revalued in this movement: the discussed qualification of what characterises a network and the frictions between the entailed ambitions in terms of maintenance and the usual accounting frameworks.

Conclusion

What we have observed in the French water systems finds echoes in different sectors and countries, in which maintenance progressively appears as a central figure of infrastructure management. The reproblematisation of maintenance that is unfolding in Northern countries does not lead to the emergence of a renewed “infrastructure ideal” in the sense of Graham and Marvin (2001), though. Rather, it reflects a shared preoccupation, a concern for infrastructure durability, which national and international organisations articulated from a specific technical, economic and environmental configuration. Addressed as interconnected issues, the material decay of infrastructures, the tightening of financial constraints and the acceleration of the climate crisis form the foundations of a specific and situated age of infrastructure maintenance (Denis and Florentin, 2024).

As we have shown in this paper, such multifaceted problematisation results in a twofold displacement. First, it reinforces and broadens what maintenance means, and how it is concretely enacted. From European institutions to local utilities, what is at stake is a redefinition of the scope of maintenance itself, which should no longer be—or never should have been—limited to daily repair operations. This is what gestion patrimoniale refers to in French, and what we propose to term “more-than-corrective maintenance”: an infrastructure management framework that encompasses preventive, corrective, and sometimes predictive maintenance. This should allow utilities to better tackle the diverse challenges to make infrastructures hold and last.

Second, we showed that, in this movement, infrastructures are not simply regarded as ageing technical assets that need to be managed more rationally. They are considered anew. Infrastructures are partially rediscovered through proximal enquiries, their life is carefully documented, and their values are rediscussed.

The reshaping of infrastructure management under the aegis of maintenance thus contributes to an ongoing and evolutive redefinition of the ontology of infrastructures, questioning the ways and possibilities to apprehend it, to know it, to manage it and to value it. Importantly enough, the more-than-corrective maintenance programmes we identified in this paper do not aim for permanence or status quo. On the contrary, they call for acknowledging infrastructures as singular and situated assemblages that evolve and change in specific manners. Embracing idiosyncrasies and uncertainty, this form of maintenance is generative (Duclos and Sanchez Criado, 2019), opening up new technical challenges, new labour activities and new organisational transformations, possibly epitomising one facet of what some researchers have called a post-networked urbanism (Rutherford and Coutard, 2016).

Obviously, the ongoing processes recounted in this paper should not be understood as a simple and unidirectional trajectory that is occurring, will occur or should occur in every infrastructural sector. Nor is maintenance to be fetishised as a necessarily ecology-driven or wellbeing-oriented practice as some maintenance scholars sometimes suggest. What the water stories we have gathered here invite us to consider is rather that an extended maintenance policy opens up opportunities to transform infrastructure management and imagine renewed locally embedded territorial strategies that will somehow resist any form of large standardisation.

This not only concerns infrastructures and the way they are practically taken care of. It also has important political implication, since it places utilities themselves and all urban operators at a crossroad, with the possibility for them to endorse new roles in the urban arena. Even though placing maintenance at the centre of infrastructure management cannot lead to radically question the way large infrastructures have been designed and constructed, we showed that more-than-corrective maintenance policies make infrastructures matter in a somewhat different way. Doing so, they may lead to destabilising former established hierarchies and participating to the reassembling of urban orders. For instance, networked infrastructures have long been downplayed by the highways and roads departments within local administration, the works made on the latter conditioning most of the interventions done on the former. The emerging problematisation of infrastructure through more-than-corrective maintenance is transforming this hierarchy, when not sometimes entirely reversing it (Denis and Florentin, 2022a): roads construction specifications has thus become in several places largely determined by the water, gas and electricity networks agenda, to ensure a proper maintenance of them. Maintenance activities offer here a site to negotiate the accuracy of former urban priorities and hierarchies and reconsider their adequation with an integrated urban management.