Interdependencies in infrastructure systems-of-systems

Infrastructure networks have been designed and evolved as sector-specific entities: electricity, gas, road, rail, digital communications and so on. They inevitably have interdependencies between each other, but these have not, until relatively recently, attracted a great deal of attention. One network operator (e.g. a water utility) has tended to assume that the networks upon which it is dependent (e.g. electric power, transport for the delivery of chemicals) will continue to function come-what-may. That optimistic assumption has been shattered by a series of damaging and disruptive incidents. Though by no means the most disruptive, the well-documented account of flooding of an electricity substation in Lancaster in December 2015 is revealing in that it demonstrates how rapidly all societal functions can cease these days without electric power for information and communications technologies. This points to a process of rapid technological convergence upon electricity as the dominant energy vector and digital connectivity being essential for all infrastructure networks to function. These are not the only interdependencies, but they are the most ubiquitous and are becoming reinforced by the dominant direction of technological change, towards electrification (in part motivated by the desire to decarbonise the power sector) and digitisation. Other significant forms of interdependence include the dependence of most thermoelectric power plants upon water for cooling, and the dependence of practically all infrastructure upon transport networks (mostly roads) to transport workers to infrastructure sites, even during catastrophic events.

Interdependencies also bring opportunities, for example by using the batteries in electric vehicles to store electricity from renewable energy supplies and better match electricity demand with supply. Risks of water shortage in many parts of the world mean that there are increasing efforts, led by countries like Singapore and Israel, to ‘close the urban water cycle’ by directly reusing ‘waste’ water in sewage for urban and agricultural water supplies. There is also growing interest in the ways in which green infrastructure (like green spaces and wetlands in urban areas) can substitute for ‘grey’ infrastructure, often by providing multiple services like sustainable urban drainage, purification of waste water and cooling urban areas (through shade and transpiration).

These interdependencies influence every-day planning for infrastructure investment and operation, and they influence infrastructure performance during extreme events. One set of crucial imponderables for energy system planners is how many electric vehicles there will be in the future, when and where they might need to be charged and whether they can be used as battery storage devices at times of excess renewable energy supply. Another set of questions relates to the growing need for digital infrastructure to service transport technologies like autonomous vehicles. Planning one infrastructure network in isolation will overlook these crucial interdependencies. The framework set out below is designed to explicitly and conveniently incorporate quantified understanding of interdependencies in a system-of-systems methodology.

A framework for infrastructure systems-of-systems analysis

The Infrastructure Transitions Research Consortium (ITRC) is a consortium of seven of the UK’s leading universities which has been working since 2011 to develop, test, demonstrate and deliver methodology for national infrastructure assessment (NIA). The ITRC was backed by £4.7 million of funding from the UK’s Engineering and Physical Sciences Research Council and in 2015 was awarded a further £5.3 million to continue the research programme through to 2020. The ITRC has adopted a ‘system-of-systems’ approach to infrastructure, promoting cutting edge research on the interdependencies between the following infrastructure sectors: energy (electricity, gas), transport (road, rail, ports, airports), digital communication (fixed, mobile, satellite), water (supply, waste water treatment, drainage, flood protection) and solid waste. The assessment methodology is depicted in the diagram shown in Figure 1.

ITRC NISMOD model diagram

Figure 1: Schematic overview of the use of NISMOD for national infrastructure assessment.

The system-of-systems framework shown in Figure 1 uses scenarios (top left box) to explore uncertainty in a range of possible futures: the global and national economy; population/demography at national and local scales; climate change; technological development. By examining a wide range of possible scenarios, we can test the sensitivity of infrastructure policies and plans to different possible future states of the world. These scenarios explore the possible range of contextual factors that are largely outside the control of decision makers responsible for national infrastructure. Of course there are feedbacks between the infrastructure system and these factors, notable in terms of regional economic development: provision of infrastructure can stimulate the economy, provide new employment opportunities and hence change where people live. These feedbacks are important to recognise but are also much more difficult to quantify – attempting to model them would introduce additional complexity that for the time being we consider to be unwarranted. We prefer to explore the possible impacts of infrastructure investment on regional economies by using scenarios that test the possibility of additional growth being stimulated by infrastructure provision.

Alongside scenarios, our framework explores one or more infrastructure strategies. Strategies are sequences of infrastructure investments and policy/regulatory interventions that are intended to modify demand for, or provision of, infrastructure services. Strategies must be sufficiently adaptable to cope with uncertain scenarios, for example of population growth. On the other hand, a national infrastructure strategy must give a clear sense of direction about what policies and investments need to be implemented, where and when. The balance between supply-side and demand-side interventions depends on the circumstances and the availability of technologies. Over-provision of new infrastructure (supply-side) can result in infrastructure being under-utilised or resources being used inefficiently. This has stimulated increasing interest in demand-side instruments like pricing (e.g. congestion charging) or regulation to mandate efficiency standards (e.g. for electric appliances). However, these are not cost-free and can require strong political commitment in the face of resistance by infrastructure users.

Scenarios and strategies provide the boundary conditions for the coupled system modelling depicted in the centre of the diagram in Figure 1. NISMOD runs in a simulation mode i.e. for a given set of input conditions (scenarios and strategies) the coupled system models simulate system performance and compute a series of metrics through the simulation period (typically decades into the future) which can be used to evaluate system performance (see below). NISMOD also keeps track of all of the significant interdependencies between infrastructure sectors. Though there are many interdependencies, from an infrastructure assessment and planning perspective, the interdependencies that matter most are when (i) demand for one infrastructure sector is highly correlated with demand for another (e.g. domestic demand for water and energy)  and/or (ii) when one infrastructure sector can potentially consume a significant proportion of the capacity of another, notably in power generation, which is responsible for 40% of non-tidal surface water abstractions. Another instance of the latter type of interdependence would become critical if there were to be largescale uptake of electric vehicles, which could eventually use more than 15% of electricity generation in 2050.