Revolutionizing water management

How will the technological advances of the fourth Industrial revolution impact how we manage water?

The rise of technologies such as smartphones and Uber are more than merely convenient. These shifts have fundamentally changed many aspects of our daily activity, from how we order food, access transportation, share information and more. We have seen this change exponentially through our lives, from Google Maps telling us the best route to avoid traffic, to Netflix suggesting the next series you may enjoy.

These technological shifts are revolutionary – so much so that they’ve been formally classified as such. We are living in the midst of the fourth industrial revolution. The first industrial revolution introduced mechanized work, powered by steam and water. The second meant mass-produced goods, powered by electricity. The third saw the rise of computer technology. The fourth brought us those smartphones and Uber.

At its core, the fourth industrial revolution is the generation of a digitized world, coupled with the power and intelligence to process this data in real time to make better decisions. Some key technologies this encompasses are:

  • Big data – the generation of massive amounts of data from sensors, satellites, drones etc.
  • Internet of Things (IoT) – a network of devices (such as sensors, pumps etc.) connected by the internet.
  • Machine learning – algorithms which aim to continually improve or optimize predictions or decisions, based upon the outcome of a sample set.
  • Artificial Intelligence (AI) – aims for computers to perform in an intelligent manner, for example in diagnosing and solving problems which they have not explicitly been designed to solve.
  • Blockchain – an open and distributed electronic ledger. Notably, it results in a transparent record and requires consensus to change any information.

The implications of these disruptive technologies are endless. When it comes to water, the fourth industrial revolution holds potential solutions to many of our concerns about the future, like population growth, climate change and environmental damage. And many of these solutions are either already here, or not that far off. I have provided a few examples of some key areas, but there are many other examples and opportunities.

Active management and early warning systems

The development of IoT sensors (whether stationary or on drones/satellites) means a constant stream of data, which can be analyzed by AI. This will allow water system performance to be continually assessed, opening the door to adaptive management and systems optimization based on real-time information on prevailing conditions.

Every year, approximately 25 per cent of Canada’s drinking water is lost to leakage from pipes – that’s equivalent to drinking water for almost nine million people. Using IoT sensors, leakage within buried pipes can be assessed and hotspots pinpointed proactively. Failure risk could be continually assessed and coupled with IoT valves to isolate burst pipes.

At some point, robotics may be able to identify issues and repair buried infrastructure. This will be a real game-changer compared to the disruption and cost associated with physically digging up pipes (frequently under roads).

Real-time decision making

As cities grow and urbanization continues, existing stormwater management infrastructure faces ever-increasing pressure – especially in older cities where combined sewers still exist, which can overflow raw sewage directly to rivers and even roads during flooding events. There are limited options for expansion, except large-scale replacements which come at high cost and heavy disruption.

As part of the Turcot Interchange in Montreal QC, WSP integrated the Interchanges’ stormwater management system with the City of Montreal’s dynamic management system. The dynamic system manages water storage, and aims to reduce existing combined stormsewer overflow to the St. Lawrence River and protect the Turcot Interchange infrastructure.

The system processes real-time weather forecasts, as well as wetland requirements and stormsewer capacity information, to determine where the water should go. The preference is to divert water to a nearby wetland, reducing pressure on the stormsewer network (and hence overflows to the river) and enhancing the wetland condition. Once the location has been determined, IoT enabled devices pump stormwater to either the wetland or stormsewer.

To deliver this innovative solution, engagement and alignment were required between several parties with different roles, responsibilities and interests – the Ministry of Transportation Quebec (roadway), the City of Montreal (stormwater management), the Ministry of Sustainable Development, Environment, Fight Against Climate Change (wetland health) and KPH Turcot (contractor).

Early warning systems

In 2018, Canada saw the fourth-highest flood damage on record, exceeding $1 billion. However, that number is predicted to increase with the impacts of climate change and increased urbanization.

Figure-EN

WSP in the UK, in collaboration with Pyterra and the University of Surrey, developed a tool which utilizes machine learning to provide near real-time flood predictions. The tool is fed with current river water level data from IoT enabled sensors, and uses weather conditions and forecasts to generate flood predictions. Since the tool utilizes relationship analysis, rather than conventional detailed hydrologic and hydraulic modelling, the results can be generated rapidly in comparison to conventional approaches. For those catchments where there were long records of good quality data, the tool achieves an average 80 per cent accuracy.

With early warnings, simple measures and adaptation strategies such as sandbags can be put in place to significantly reduce damage and disruption from impending floods.

The second stage of this study incorporated active management of up-stream storage areas. The aim was to provide more flood storage in upper catchments of a river to reduce downstream flooding. This was provided by taking the impending flood predictions and draining upper catchment flood storage areas (both natural and man-made, such as reservoirs) before the storm arrives and without causing downstream flooding. The result is more storage available during the storm, and hence less downstream flooding.

Re-writing the guidelines

Many engineering principles governing design are established in extensive and sometimes prescriptive design guides. These design guides provide summary categories, best practices, established thresholds and regional values, based on decades of practical experience and summary of incomplete knowledge or data. These can lead to ingrained, siloed and process-driven design engineering, and significant missed opportunities for a subject such as water which crosses so many individual disciplines.

This approach will be surpassed by big data-driven and AI-aided design, accounting for local variations whilst balancing and integrating with surrounding disciplines. Just as we saw the AI at Facebook develop its own language because English was too inefficient, AI will also help us see opportunities in engineering.

Drainage design

Conventional drainage design generally seeks to mitigate flooding up to a design storm limit, whilst removing the worst of sediment. Typically, systems are designed conservatively and involve minimal moving parts, due to the risk of poor maintenance and subsequent failure. As a result, diffuse systems such as rain barrels or cisterns are not often incorporated into flood mitigation solutions – how can you guarantee they will be empty or working during a storm? Following construction, the actual system performance is difficult to determine.

With the integration of IoT systems and AI, actual system performance and functionality can be assessed continuously and more accurately. If something is not working, it can be diagnosed and maintenance requested automatically. This means that these diffuse systems can now be relied upon and offer different solutions to the traditional larger pipes or bigger storage areas.

IoT sensors and AI will help enable adaptive system management, which – when combined with earlier and better warning systems (as discussed above) – will enable municipalities to operate systems differently under normal and flood conditions. Impacts of new drainage branches can be assessed across a whole system (meaning piped and natural systems), with the new focus on the whole system performance, not the singular additional branch.

This design approach will plan for extreme events, to help ensure damage from exceedance is not catastrophic. In some cases, city planners have started to integrate temporary flood storage into public space to relieve pressure on the creaking sewer systems, such as Copenhagen’s floodable parks or Rotterdam’s Blenthemplein water square.

Decentralization and integration of systems

The concept of decentralization for water is similar to decentralization of energy. We have heard about buildings setting up solar panels and trading energy with one another through blockchain, or installing batteries so they can be self-sufficient even with no sunshine. The same concept can be applied to water. IoT-enabled blue roofs could collect water, which can be used within the building as grey water, reducing pressure on storm sewers and potable water supplies during critical periods and reducing building heat loss.

While the focus of decentralized systems is predominately in highly-urbanized areas where increased water supply and stormwater management capacity is expensive and disruptive, it can also be important in remote areas (such as Indigenous communities) where centralized infrastructure does not exist.

City of Altamonte

It is ironic that we have two completely separate systems: one designed to get rainfall away from homes, often ultimately discharging to a river, and another which abstracts water from a river, treats it and pumps it back to the homes as drinking water.

The City of Altamonte developed a system to treat drainage water back to drinking water quality standards in an energy-efficient manner, which does not produce a waste brine. The water can then be provided back to homes. In this way, the stormwater system can be transformed into an efficient water collection system.

Due to risks of contamination, this type of system can only be possible with advanced treatment combined with the protection afforded by continuous real-time water quality monitoring and AI.

Data coverage and access

The fourth industrial revolution hinges on reliable, interoperable and accessible data. Data availability (and quality) is a significant current challenge in water management, especially in remote locations. Open Data is making data much more accessible, and when coupled with blockchain technology, it can significantly improve reliability and interoperability.

Satellite imagery as well as drones and autonomous vehicles are rapidly shrinking the geographic distances and boundaries for data collection. Projects like Microsoft’s AI for Earth, World Resource Institute’s Aquaduct water risk atlas and global flood analyzer are continually providing more and better information. The continued improvement in this area is helping implement sustainable water management practices in remote locations.

Barriers to adoption

To maximize the opportunities that the fourth industrial revolution brings, we will need to overcome several significant barriers. These could warrant a whole article to themselves, but include things such as:

  • Regulation – the water industry is tightly regulated and generally risk-averse, as consequences of poor management can include flooding and polluted drinking water. This makes decision-makers reluctant to adopt new and relatively unproven solutions.
  • Roles, responsibilities and ownership – in my opinion, the biggest current barrier. IoT-connected rain barrels on everyone’s homes soaking up stormwater would be great – but who pays for them, who maintains them and who benefits from their presence? Rarely do all three of these things fall to the same stakeholder, and in many cases, this will require public-private collaboration to maximize potential. To a certain extent, maintenance concerns could be overcome in areas of stormwater management charging, with rebates or incentives being linked to IoT feedback.
  • Energy consumption – the energy required to power sensors and systems will be significant. For example, BitCoin’s energy consumption last year was reportedly higher than Ireland’s.

Conclusion

The fourth industrial revolution is progressing at an exponential rate. Some of the things I’ve discussed here are possible today, and some will be soon. Personally, the most exciting aspects are not just the technologies themselves, but the changes which they enable. Instead of designing systems piecemeal and to cope with defined limits, we will be able to design and proactively manage an integrated and flexible system which can adapt in real time to prevailing conditions to manage floods, environmental requirements and water scarcity more efficiently. These systems can make the most of water resources when they are scarce, and help provide resilient solutions to the extensive stormwater and river flooding experienced across Canada.

However, to harness the full potential of these opportunities and benefits, we not only need to embrace the technological advances, but we must also work together (public and private) to change our relationship with water. We must integrate historically siloed disciplines and systems, and we must understand we cannot solve flooding (or droughts) – but we can manage it responsibly in a way which minimizes damage and disruption.


More on this subject



Related themes and tags

Places

Resilience

Society

Technology