Board guides on connectivity for continuous water monitoring

Water companies measure water quality and network conditions at sites across their catchments, treatment works, distribution networks and rivers. Where monitoring supports a remote alarm, corrective action or central record, each reading must reach the system responsible for that function.

Two CSL guides set out two key decisions for water companies: which monitored assets require connectivity through two independent operator cores, and how publication within the hour should be defined and evidenced.

The drinking water guide covers supply in England and Wales, and the receiving water guide addresses Defra’s continuous water quality monitoring programme in England. Each examines the requirements and guidance published as at 4 November 2026 and sets out a specific connectivity decision for the water company. The guides explain when that decision is needed within a programme and what supporting evidence should be agreed with suppliers.

Both guides are also available at the end of this page.

In drinking water supply, an online reading can do more than record a result. The Drinking Water Inspectorate’s Drinking Water Safety Plan guidance covers the assessment and management of risk from catchment to consumer. It states that monitoring linked to an automated corrective action could be considered a control measure: a step that prevents or reduces a risk to drinking water. Where a water company’s risk assessment report identifies a significant risk, the Water Supply (Water Quality) Regulations 2016 for England and the separate 2018 Regulations for Wales allow the Secretary of State or the Welsh Ministers, respectively, to require specified measures through a regulation 28(4) notice. Published notices can require online monitoring with alarms and fail-safe shutdown, and in 2025 the Inspectorate issued 87 regulation 28(4) notices in England. Where the system that acts on a reading is remote from the measuring instrument, the action also depends on the connection that takes the reading to it.

As traditional phone lines move to digital by January 2027 and 2G services are switched off from 2029, monitoring points that rely on an analogue PSTN line or a 2G-only device will need an alternative connection. This provides a natural opportunity to decide which mobile networks each device can use and which monitored assets need connectivity through two independent operator cores. An operator’s core authenticates devices using its subscriptions and routes their traffic. A fault affecting that core can therefore disrupt multiple devices, potentially across an entire monitored estate, even where they connect through different available radio networks. The guide explains how water companies can use the replacement programme to classify their instrumented estate. Connectivity can then be specified according to each instrument’s regulatory role, site conditions, reporting interval and the consequences of losing its reporting path. A second operator profile on an independent core can be specified when a device is replaced or reconfigured. Adding it later may require another site visit if the device cannot be updated remotely.

Where a reporting path fails, a reading can arrive late, delaying corrective action and potentially extending the time before normal supplies are restored. If this increases the duration of an event assessed by the Inspectorate, the event can receive a higher Event Risk Index score. The index includes an impact score calculated by multiplying the maximum population affected by the duration in hours.

Defra’s provisional technical guidance states that water quality data must be made publicly available in near real time, within one hour. Section 5.3 of the assurance standard appended to Defra’s interim technical standard repeats this publication requirement. Neither passage explicitly defines the starting point for the hour, so the CSL guide uses the measurement timestamp. Readings must normally be taken at least hourly, increasing to at least every 15 minutes while a storm overflow is operating or final effluent quality is, or could be, affected, and for 24 hours after the triggering event ends.

Before it reaches the public, each reading passes through five parts: the sonde that measures in the river, the telemetry outstation, the communications link, the collection systems and the publication platform. Together, these form the reporting path. Depending on the delivery model, its constituent parts may be provided by one supplier or divided among several. If the one-hour requirement is divided into separate specifications for each part, every supplier could meet its own specification while the reading still reaches the public too late.

Section 4.3 of the assurance standard appended to Defra’s interim technical standard concerns sensor faults at the sonde. It states that equipment downtime due to those faults will not exceed 10 per cent of the monitoring time each year, with no data gaps longer than two consecutive weeks. The standard also specifies outstation buffering and data-centre requirements, but does not set an availability figure for the reporting path as a whole. The guide therefore recommends that the water company sets an availability outcome for the complete reporting path, such as the proportion of readings that must be made public within one hour of measurement.

The guide gives CSL’s advice in two parts. The water company sets publication outcomes for the complete reporting path, and agrees the evidence for each part with the organisation that supplies it. These outcomes and the evidence required should be agreed during procurement, confirmed in the contract and checked at design approval and acceptance. Factory and site acceptance tests show when each reading was measured and published under the agreed test conditions. Performance measured in service shows the reporting path’s availability across a year.

For a wider explanation of the regulatory background, read Designing Resilient Connectivity for Continuous Water Monitoring. The article covers the requirements behind both guides and sets out seven provisions CSL recommends including in a resilient connectivity specification.

CSL provides the communications link as a managed IoT connectivity service, from general monitoring connections to architectures requiring independent mobile-core paths or connectivity at sites with power, coverage or installation constraints.

CSL DualCore® provides two secure connectivity paths through independent mobile operator cores, so the monitored site does not depend on a single operator core. CSL delivers this architecture as a managed service through rSIM® technology or a CSL DualSIM® Router configured for core independence. With rSIM technology, the SIM tests the active path and can switch profile if the test fails, without relying on the connected device to make the switching decision. Where mobile signal is weak at a monitoring site, CSL Outpost can combine up to four active mobile connections, using appropriately positioned external antennas to make the most of the available coverage. The mix of networks and connections is configured for the site. Where usable mobile coverage is unavailable, CSL Satellite, or CSL Outpost incorporating a satellite path, can provide connectivity, subject to suitable power, installation conditions and a clear view of the sky. To discuss the connectivity requirements for a monitoring estate or programme, contact sales@csl-group.com.

• Continuous Monitoring in Drinking Water Supply
  A board guide in the Critical Connectivity® Research Series. First edition, November 2026. 11-page PDF.
• Continuous Water Quality Monitoring of Rivers
  A board guide in the Critical Connectivity® Research Series. First edition, November 2026. 11-page PDF.