--- title: "Backup power at critical sites | Nexus Observatory" description: "When the grid fails, hospitals, water plants, emergency centres and telecom hubs run on generators and batteries. Many fail to start…" url: https://observatory.global/industries/resilience/backup-power/ publisher: Nexus Observatory Inc. updated: 2026-11-02 --- [Nexus Observatory](https://observatory.global/) / [Civil resilience](https://observatory.global/industries/resilience/) / [Specialized fields](https://observatory.global/industries/resilience/fields/) / **Backup power at critical sites** # Backup power that starts when it is needed. When the grid fails, hospitals, water plants, emergency centres and telecom hubs run on generators and batteries. Many fail to start, run out of fuel or trip under load. Monitoring every generator, tank, battery and transfer switch, with tests on record, turns backup power from a hope into a fact. [Plan a deployment](https://observatory.global/contact/deployment/) How it works [Station Builder](https://observatory.global/studio/builder/) **Industry**: Civil resilience **Measurements**: 5 **Sources cited**: 2 1. Civil resilience · Backup power at critical sites The outage is the test When the grid fails, hospitals, water plants, emergency centres and telecom hubs run on generators and batteries. Too often a generator does not start after months idle, a tank is low, or a battery no longer holds. Standards ask for regular tests; between them, backup power is usually watched by no one. 2. What is measured Engine, fuel, battery and switch Meters read generator output and load, sensors read fuel level and quality in day and bulk tanks, battery systems report state and runtime, contacts read the transfer switch position, and engine sensors read vibration and temperature during every run and test. InstrumentsPower meters · Fuel sensors · Battery monitors · Switch contacts · Engine sensors 3. The base node Runtime, computed The node computes how long each site can run from fuel, load and battery state, detects failed starts and failed transfers, flags fuel quality problems and engine faults, and records every test and outage automatically, so the evidence inspectors ask for is complete. 4. Radio and links Reporting through the blackout Cellular carries power state from each site and survives grid outages where cell sites have their own backup. Sites that must report through long regional blackouts add a satellite path. The monitor runs on the site's protected supply, so it reports the outage it is there to watch. LinksPublic cellular · Satellite 5. How it works From a blackout to a fuel truck A storm blacks out a region. Generators start at forty sites; one fails, and its crew is sent at once. The emergency centre sees runtime at every site, and fuel trucks go first to the water plant and the hospital with the shortest time left. 6. From order to operation · Order Scope the critical load The scope reads each site's critical loads, its generators, tanks, batteries and switches, test records, outage history and the fuel contracts. Nexus Observatory Inc. returns the monitoring kit per site, the runtime and fuel model, test reporting and a price per site and per year. [Plan a deployment](https://observatory.global/contact/deployment/) 7. From order to operation · Survey Survey every power room Crews survey generator rooms, tank farms, battery rooms and switchboards with the site engineers: controller interfaces, tank access for sensors, meter points, switch contacts and backhaul coverage. Each finding is recorded with photographs and single-line drawings. Findings shape the kit for each site before anything is built. 8. From order to operation · Permits Electrical safety and permits Work on switchboards and generator controllers follows the site's electrical safety rules and permits to work, often in planned windows for hospitals and plants. Fuel system work follows the fire safety rules for flammable liquids. Every permit is kept with the site record. 9. From order to operation · Build Build to the site's single line Kits are built to each site's single-line diagram: meters, fuel sensors, contacts and engine sensors mapped to the gateway, nodes enrolled with their identities and configured with tank sizes and loads. Everything is tested together before shipping. Nothing ships until the whole kit has passed its tests together. [Open the Station Builder](https://observatory.global/studio/builder/) 10. From order to operation · Install Install in planned windows Crews fit meters, fuel sensors, transfer switch contacts and engine sensors during planned windows, with critical loads protected throughout. Gateways and nodes are mounted in the power room, backhaul is connected and every point is photographed. Crews follow the site safety plan throughout the work. 11. From order to operation · Commission Commission with a load test Each node proves its identity. A planned load test starts each generator, transfers the load, runs it and transfers back, while every reading is checked against the controller and runtime is compared with fuel actually used. Results go into the test record. 12. From order to operation · Operate Know the runtime Operators see every site's generators, fuel, batteries and runtime. In an outage, fuel deliveries go first to the sites with the shortest runtime, failed starts send crews at once, and every test and outage is recorded for inspectors and regulators. 13. From order to operation · Maintain Test on schedule, keep the record Scheduled generator tests and battery checks are recorded automatically, fuel is sampled and polished when quality drops, engine trends schedule servicing, and node software updates roll out in rings. Every test and repair is kept with the site. Maintenance history stays with each site for its whole life. [Book field services](https://observatory.global/contact/field-service/) 14. Data products For managers, crews and suppliers Emergency managers and site operators receive runtime at every site; maintenance crews receive failed start and transfer alarms; fuel suppliers receive fuel needs by site; inspectors, auditors and regulators receive test and outage records, each tied to the site, the generator and the time. [See the data products](https://observatory.global/data-products/) 15. Evidence Tests and outages on record Monitoring and test records follow the emergency and standby power standard, generators are assessed against the generating set series, and records support the resilience duties of critical entities. Every start, transfer, test and outage is signed and kept, ready for inspection. Drawn againstNFPA 110 · ISO 8528 · Directive (EU) 2022/2557 16. What you get Microgrids and off-grid site power PV, storage and generator sizing for instruments and edge sites, with telemetry and remote control. Delivered as hardware + software and managed service, for remote sites, telecom, agencies, communities. Scale: Per site; fleets. Service levels are written as numbers you can measure on your own dashboard. [Talk to sales](https://observatory.global/contact/sales-enquiry/) 17. Next step Plan backup power at critical sites with us Tell us the site, the measurements and the rules it answers to. We reply with a survey proposal, a design and a scope for backup power at critical sites. An engineer replies with a survey proposal and a scope, and the plan grows from one site to the network you need, at your pace. [Plan a deployment](https://observatory.global/contact/deployment/) [Open the Station Builder](https://observatory.global/studio/builder/) [Open the Studio](https://observatory.global/studio/) **01** The place and its hazards ## Where it runs, and what can go wrong. Operating contextEmergency and standby power systems are designed, tested and maintained to their own standards, generating sets are rated to the international generator series, and operators of critical entities must show their services survive disruption. Yet backup power is often tested rarely and watched by no one between tests. Hazards - Generators that fail to start after months idle - Fuel that runs out, degrades or is stolen - Batteries that no longer hold their rated time - Transfer switches that fail to transfer - Long outages that outlast the fuel on site **02** What must be measured ## What, where, how often, and under which rules. Each measurement names where it is taken, how often, and the public standard or regulation behind it. Numbers in the sources column open the cited source below. | Measurement | Where | Cadence | Sources | | --- | --- | --- | --- | | Generator start, run and load | Each generator's controller | Continuous, events on start and stop | 12 | | Fuel level and quality | Day tanks and bulk tanks | Continuous | 1 | | Battery state and runtime | Uninterruptible power and DC battery systems | Continuous | 1 | | Transfer switch state and transfer time | Every automatic transfer switch | On event and in tests | 1 | | Engine vibration and temperature | Generator engine and alternator | During runs | 2 | **03** The observation stack ## From one base unit to a network. The same blocks at every scale: instruments and a base node, a site with its radio, a fleet run as one, and a network federated on the open protocol. 1. **01** Base unit A power point: meters, fuel sensors, transfer switch contacts and engine sensors on an industrial gateway with a node. [Build a station →](https://observatory.global/studio/builder/) 2. **02** Site A site: every generator, tank, battery and transfer switch of one hospital, plant or hub, with its own backhaul. [Plan a site →](https://observatory.global/contact/deployment/) 3. **03** Fleet Every critical site of an operator or a city watched as one fleet, with fuel and runtime across all of them. [Managed offers →](https://observatory.global/industries/resilience/solutions/) 4. **04** Network Emergency managers and fuel suppliers share runtime and fuel needs across operators through open standards. [The open protocol →](https://observatory.global/protocol/) - **01** Energy and power-quality meter - **02** Vibration and condition monitor - **03** OT gateway (OPC UA, Modbus) - **04** Edge node (Wedge) - **05** Cellular modem/router - **06** Generator - **07** Battery case Radio and links ### Public cellular Cellular carries power state from each site; it survives grid outages where cell sites have their own backup. Radio and links ### Satellite Sites that must report through long regional outages add a satellite path. Edge computeThe node computes runtime from fuel, load and battery state, detects failed starts and transfers, flags fuel quality problems and engine faults during runs, and records every test and outage. PowerThe monitoring runs on the site's protected supply with its own battery, so it reports the outage that it is there to watch. **04** How it works ## From the instrument to the product. One observation's path, tier by tier, with the interface on every link and a bill of materials below; then how the one node grows to a site, a fleet and federated networks. Backup power at critical sites: data path – of 07 01 Instruments Measurement with sensor identity andcalibration 02 Gateway 03 Node – Wrap in an observation envelope with timequality– Measure quality and uncertainty, attachquality flags– Buffer locally (store-and-forward when thelink drops) 04 Backhaul Commit with transactional outbox andpublish, delivered at least once 05 Site Consumer inbox deduplicates and applieseach record once into governed data 06 Data store Governed product through API, map orexport, correction-aware 07 Products ModbusOPC UAMQTT 08 Power and structure Supplies 12 V DC, 24 V DC, 48 V DC to thefield equipment 01 Energy andpower-quality meter SDI-12 · RS-485 · Ethernet 02 Vibration and conditionmonitor Analogue I/O · RS-485 ·Ethernet 03 Generator RS-485 · Ethernet Fuel level and quality sensorsin day and bulk tanks Not in the reference kit Transfer switch positioncontacts Not in the reference kit 04 OT gateway (OPC UA, Modbus) Ethernet · RS-485 · CAN · AnalogueI/O 05 Edge node (Wedge) Ethernet · Fibre · USB · RS-485 ·BMC 06 Cellular modem/router Ethernet · USB · RS-232 Site MQTT 5 with Sparkplug B Governed data store Versioned, correction-aware Runtime at every site Failed start and transfer alarms Fuel needs by site Test and outage records 07 Battery case USB RS-485 · Modbus RS-485 · Modbus RS-485 · Modbus Analogue I/O Analogue I/O Ethernet · opc-ua Ethernet LTE/NR From the instrument to the published product: each band is a tier, each connector names its interface, and the numbers match the bill of materials. Hover or focus a block to trace its path; scroll or step to follow one observation. | Item | Equipment | Data ports | Protocols | Power in | | --- | --- | --- | --- | --- | | 01 | **Energy and power-quality meter** Meter for voltage, current, power, energy and power quality at a site or on a feeder. | SDI-12RS-485Ethernet | Modbus, IEC 61850 (MMS, GOOSE), DNP3 (IEEE 1815), MQTT 5 | 12 V DC | | 02 | **Vibration and condition monitor** Accelerometer-based vibration and temperature sensor for rotating machinery and structural condition monitoring. | Analogue I/ORS-485Ethernet | Modbus, OPC UA, MQTT 5 | 12 V DC | | 03 | **Generator** Fuel generator with automatic start for backup or primary site power. | RS-485Ethernet | Modbus | Fuel | | 04 | **OT gateway (OPC UA, Modbus)** Read-only bridge from industrial fieldbuses and controllers (Modbus, OPC UA, DNP3, IEC 61850, CAN) to the site network, behind a security boundary. | EthernetRS-485CANAnalogue I/O | OPC UA, Modbus, DNP3 (IEEE 1815), IEC 61850 (MMS, GOOSE), MQTT 5 | 24 V DC | | 05 | **Edge node (Wedge)** Rugged edge node that boots from signed media, attests itself and establishes a site cell from little or no infrastructure. | EthernetFibreUSBRS-485BMC | MQTT 5, DMTF Redfish, IEEE 1588 PTP, NTP with NTS, OPC UA | 12 V DC, 24 V DC, 48 V DC, AC mains | | 06 | **Cellular modem/router** LTE or NR modem or router that uses a public carrier or private network as backhaul. | EthernetUSBRS-232 | 3GPP E-UTRA air interface, 3GPP NR air interface, OMA LwM2M | 12 V DC, PoE | | 07 | **Battery case** Portable battery pack with DC outputs for field kits. | USB | Modbus | AC mains, HV DC, USB PD | How it scales 01 Instruments Arbitrary qualified sensors and instruments whosemeasurements are ingested with calibration, units,geometry, time, custody and uncertainty. ×n 02 Node The minimum useful deployable unit: server, ruggedappliance, radio node, sensor gateway, satelliteterminal, AI, scientific, vehicle or virtual node. Node Node agent 03 Site The minimum independently operable localgovernance and failure domain: one or more nodesplus local network, identity, policy, evidence,storage, products and operator surfaces. SITE Node Node Node ×n 04 Cluster An optional scale or specialisation domain forscheduling, shared storage, accelerators or datalocality. CLUSTER ×n 05 Fleet Coordinates plans, acknowledgements,convergence evidence, update campaigns andcorrection propagation across nodes and sites. FLEET Fleet controller ×n 06 Federation An exchange of products, capabilities, events orwork across approved, independently administeredtrust domains, using explicit trust, policy,data-locality, version and correction contracts. TRUST DOMAIN TRUST DOMAIN TRUST DOMAIN ×n Verifier Applications Data in the field Telemetry Control Data upstream Control Data upstream Control Exchange Evidence Events Exchange The smallest deployable unit is one node; the same unit is replicated to meet the demand. Nodes form a site, sites may form a cluster, sites and clusters are coordinated as a fleet, and fleets exchange across independently administered trust domains. Each plane is a separate channel with its own open binding. **Data in the field**: MQTT 5 with Sparkplug B; connectors for LoRaWAN, OPC UA and instrument buses. **Telemetry**: OpenTelemetry OTLP. **Control**: TLS 1.3 with mutual authentication; protobuf messages over gRPC or HTTP/2. **Data upstream**: At-least-once publication with repeat-safe keys over the control channel or NATS JetStream; QUIC MAY be used on lossy links. **Evidence**: DSSE-signed in-toto statements; SLSA provenance; SCITT transparency receipts. **Events**: Event envelopes mapped to CloudEvents 1.0; channels described in AsyncAPI 3. **Exchange**: OpenAPI 3.1; the OGC API family; SensorThings; STAC; WIS 2.0. **05** From order to operation ## Every step, from the first call to a running site. The same crew and the same record carry the work from scope to maintenance. 1. 01 Order Scope the critical load The scope reads each site's critical loads, its generators, tanks, batteries and switches, test records, outage history and the fuel contracts. Nexus Observatory Inc. returns the monitoring kit per site, the runtime and fuel model, test reporting and a price per site and per year. 1 2. 02 Survey Survey every power room Crews survey generator rooms, tank farms, battery rooms and switchboards with the site engineers: controller interfaces, tank access for sensors, meter points, switch contacts and backhaul coverage. Each finding is recorded with photographs and single-line drawings. Findings shape the kit for each site before anything is built. 3. 03 Permits Electrical safety and permits Work on switchboards and generator controllers follows the site's electrical safety rules and permits to work, often in planned windows for hospitals and plants. Fuel system work follows the fire safety rules for flammable liquids. Every permit is kept with the site record. 4. 04 Build Build to the site's single line Kits are built to each site's single-line diagram: meters, fuel sensors, contacts and engine sensors mapped to the gateway, nodes enrolled with their identities and configured with tank sizes and loads. Everything is tested together before shipping. Nothing ships until the whole kit has passed its tests together. 5. 05 Install Install in planned windows Crews fit meters, fuel sensors, transfer switch contacts and engine sensors during planned windows, with critical loads protected throughout. Gateways and nodes are mounted in the power room, backhaul is connected and every point is photographed. Crews follow the site safety plan throughout the work. 6. 06 Commission Commission with a load test Each node proves its identity. A planned load test starts each generator, transfers the load, runs it and transfers back, while every reading is checked against the controller and runtime is compared with fuel actually used. Results go into the test record. 12 7. 07 Operate Know the runtime Operators see every site's generators, fuel, batteries and runtime. In an outage, fuel deliveries go first to the sites with the shortest runtime, failed starts send crews at once, and every test and outage is recorded for inspectors and regulators. 1 8. 08 Maintain Test on schedule, keep the record Scheduled generator tests and battery checks are recorded automatically, fuel is sampled and polished when quality drops, engine trends schedule servicing, and node software updates roll out in rings. Every test and repair is kept with the site. Maintenance history stays with each site for its whole life. 1 **06** Data products and integration ## What it delivers, to whom, and into which systems. Data products - **Runtime at every site** Emergency managers and site operators - **Failed start and transfer alarms** Maintenance crews - **Fuel needs by site** Fuel suppliers and logistics - **Test and outage records** Inspectors, auditors and regulators Integration - **Building management** Power state reaches the site's building management system through its own interfaces. - **Maintenance** Failed starts and engine trends open work orders. - **Emergency management** Runtime and fuel needs are shared with the emergency operations centre during regional outages. Exchanged through Modbus, OPC UA, MQTT. **07** Questions buyers ask ## Before you scope it. Does it replace generator tests? No. It records every test and watches between them. Can it see fuel theft? Sudden level drops without running are flagged. Will it work in a regional blackout? It runs on protected power, and satellite can be added. Which generators does it support? Most controllers expose their data; others are read through meters and contacts. **08** Resources ## The sources behind this page. Public standards, regulations and guidance, cited where they apply. Each applies only where its jurisdiction or scope does. 1. **1** [NFPA 110: Standard for Emergency and Standby Power Systems](https://www.nfpa.org/) National Fire Protection Association 2. **2** [ISO 8528: Reciprocating internal combustion engine driven alternating current generating sets](https://www.iso.org/) International Organization for Standardization **09** Related ## Use cases and solutions. [Use case Remote microgrid observatorySolar, storage and power-quality telemetry with a weather station and a small cell, for a community that is reducing diesel.The full use case →](https://observatory.global/use-cases/remote-microgrid/) [Solution Microgrids and off-grid site powerPV, storage and generator sizing for instruments and edge sites, with telemetry and remote control.Solutions for civil resilience →](https://observatory.global/industries/resilience/solutions/#site-power) [Solution Structural and ground-movement monitoringGNSS, InSAR, tilt, strain and fibre sensing on bridges, dams, slopes, tailings and pipelines, with alert thresholds set by the owner.Solutions for civil resilience →](https://observatory.global/industries/resilience/solutions/#infrastructure-monitoring) [Solution Power-grid edge and right-of-way monitoringLine temperature and sag, vegetation encroachment from EO and LiDAR, substation sensors and fire-weather at the right-of-way.Solutions for civil resilience →](https://observatory.global/industries/resilience/solutions/#grid-edge-monitoring) **10** Civil resilience: specialized fields ## Other fields of this industry. [Specialized field Critical site securityPerimeter, intrusion and operational-technology monitoring for substations, pumping stations and other unstaffed critical sites, so operators see a threat to the site and its control systems at once.The full field →](https://observatory.global/industries/resilience/site-security/) [Specialized field Timing integrityTime references with holdover and monitoring for grids, telecom and finance, so a country's critical systems keep accurate time when satellite signals are jammed, spoofed or lost.The full field →](https://observatory.global/industries/resilience/timing-integrity/) [Specialized field Northern and Arctic observingWeather, permafrost, ground movement and river and lake ice stations for northern civil authorities, built for polar night, deep cold and satellite-only backhaul.The full field →](https://observatory.global/industries/resilience/arctic-observing/) ### Bring us the site. Plan a deployment, configure a station, model it in the Studio or talk to sales. The form names who receives your details. [Plan a deployment](https://observatory.global/contact/deployment/) [Station Builder](https://observatory.global/studio/builder/) [Open the Studio](https://observatory.global/studio/) [Talk to sales](https://observatory.global/contact/sales-enquiry/) [Request a quote](https://observatory.global/contact/hardware-quote/) [Nexus Observatory](https://observatory.global/)**Civil resilience** . Infrastructure that reports before it fails.[Talk to us](https://observatory.global/contact/) Civil resilience - [Overview](https://observatory.global/industries/resilience/) - [Explainer](https://observatory.global/industries/resilience/explainer/) - [Specialized fields](https://observatory.global/industries/resilience/fields/) - [Use cases](https://observatory.global/industries/resilience/use-cases/) - [Solutions](https://observatory.global/industries/resilience/solutions/) - [Platforms](https://observatory.global/industries/resilience/platforms/) - [Resources](https://observatory.global/industries/resilience/resources/) - [Contact](https://observatory.global/industries/resilience/contact/) Platforms - [Observatory Protocol](https://observatory.global/protocol/) - [Protocol & Grid](https://observatory.global/grid-services/) - [Software](https://observatory.global/software/) - [Hardware & power](https://observatory.global/hardware/) - [Sensors & instruments](https://observatory.global/sensors/) - [Connectivity](https://observatory.global/connectivity/) - [Edge compute & AI](https://observatory.global/edge-ai/) - [Earth observation & GeoAI](https://observatory.global/earth-observation/) - [Space and satellite systems](https://observatory.global/space/) - [Digital twins and simulation](https://observatory.global/digital-twins/) - [Data & licensing](https://observatory.global/data-products/) - [Field services](https://observatory.global/field-services/) - [Frontier labs](https://observatory.global/labs/) Solutions - [Early warning](https://observatory.global/use-cases/street-flood/) - [Wildfire and fire weather](https://observatory.global/use-cases/wildfire-camera/) - [Environmental compliance](https://observatory.global/use-cases/methane-fenceline/) - [Asset monitoring](https://observatory.global/industries/infrastructure/) - [Private networks](https://observatory.global/connectivity/private-5g/) - [Remote power and sites](https://observatory.global/use-cases/remote-microgrid/) - [Climate and resilience](https://observatory.global/industries/climate/) - [Site twins](https://observatory.global/digital-twins/) - [Data licensing](https://observatory.global/data-products/) Industries - [Public safety](https://observatory.global/industries/public-safety/) - [Government](https://observatory.global/industries/government/) - [Cities](https://observatory.global/industries/cities/) - [Civil resilience](https://observatory.global/industries/resilience/) - [Communities](https://observatory.global/industries/communities/) - [Humanitarian](https://observatory.global/industries/humanitarian/) - [Climate](https://observatory.global/industries/climate/) - [Water](https://observatory.global/industries/water/) - [Nature](https://observatory.global/industries/nature/) - [Oceans](https://observatory.global/industries/oceans/) - [Agriculture](https://observatory.global/industries/agriculture/) - [Energy](https://observatory.global/industries/energy/) - [Telecom](https://observatory.global/industries/telecom/) - [Transport](https://observatory.global/industries/transport/) - [Infrastructure](https://observatory.global/industries/infrastructure/) - [Mining](https://observatory.global/industries/mining/) - [Industry](https://observatory.global/industries/industry/) - [Research](https://observatory.global/industries/research/) - [Education](https://observatory.global/industries/education/) - [Health](https://observatory.global/industries/health/) - [Insurance and finance](https://observatory.global/industries/insurance/) - [Events](https://observatory.global/industries/events/) Products - [Field Deployment Studio](https://observatory.global/studio/) - [Station Builder](https://observatory.global/studio/builder/) - [Field services](https://observatory.global/field-services/) - [Data products](https://observatory.global/data-products/) - [APIs and SDKs](https://observatory.global/protocol/developers/) - [Edge runtime and fleet](https://observatory.global/software/) - [Hardware catalogue](https://observatory.global/hardware/) - [Reference designs](https://observatory.global/protocol/reference-designs/) Resources - [Learn](https://observatory.global/learn/) - [Stories](https://observatory.global/stories/) - [Environments](https://observatory.global/environments/) - [Network scenes](https://observatory.global/environments/networks/) - [Observatories atlas](https://observatory.global/observatories/) - [Earth and maps](https://observatory.global/earth/) - [Technologies](https://observatory.global/technologies/) - [Countries](https://observatory.global/countries/) - [Spectrum](https://observatory.global/spectrum/) - [Specification](https://observatory.global/protocol/specification/) - [Conformance](https://observatory.global/protocol/conformance/) - [Open source](https://observatory.global/open/) - [Data credits](https://observatory.global/credits/) Company - [About](https://observatory.global/about/) - [Open governance](https://observatory.global/consortium/) - [Security and trust](https://observatory.global/security/) - [Contact sales](https://observatory.global/contact/) - [Privacy requests](https://observatory.global/privacy-request/) © 2026 Nexus Observatory Inc. [Security and trust](https://observatory.global/security/) [Privacy choices](https://observatory.global/privacy-choices/) [Accessibility](https://observatory.global/legal/accessibility/) [Privacy](https://observatory.global/legal/privacy/) [Terms](https://observatory.global/legal/terms/)