A remote terminal unit (RTU) is the field edge of a SCADA system: installed in a distribution substation, a pump cabinet, a recloser controller or an unmanned valve site, it scans analogue and digital channels continuously, speaks IEC 60870-5-101 over serial links, IEC 60870-5-104 over TCP/IP (its registered port is 2404), DNP3, Modbus or IEC 61850 to the master station, and must keep doing so through brown-outs, feeder faults and total loss of the auxiliary supply. What looks like a quiet box in a cabinet actually carries a deceptively layered load: a small but never-zero quiescent current for scanning and memory, periodic communication bursts over radio, cellular or wired links, short high-current peaks when a modem transmits, and - critically - a defined last-gasp action in the instant the supply fails, during which the RTU must save its context, time-stamp the outage and send one final message. This first paper on nickel-metal hydride back-up for RTUs dissects that load profile, explains why back-up time on these sites is fundamentally a data-integrity and telecom-availability problem rather than a simple ampere-hour calculation, and shows where sealed, wide-temperature, low-maintenance NiMH packs fit inside a modern control cabinet.
At rest an RTU is never truly off. Its microcontroller scans digital inputs and analogue channels on a fixed cycle, runs local logic and deadband calculations, refreshes watchdog timers, keeps a real-time clock and volatile event memory alive, and maintains its protocol stacks ready to answer a master poll. IEC 60870-5-101 governs this exchange over RS-485 or RS-232 serial links common on legacy substations, while IEC 60870-5-104 carries the same application layer over TCP/IP on port 2404; DNP3 organises data into static class 0 data and event classes 1, 2 and 3, and Modbus and IEC 61850 appear alongside them on modern multi-protocol RTUs.
The consequence for power is a continuous baseline load measured in watts or fractions of an ampere, punctuated by scheduled and event-driven communication. A typical cabinet controller draws a few watts for its logic and I/O; an industrial cellular or serial radio adds a receive/standby load and then a transmit peak that can be several times the baseline for fractions of a second. Back-up sizing therefore has to cover both the long, shallow quiescent tail and these repeated peaks, which is why a back-up source chosen on average current alone can still drop the rail during a modem transmission.
Field sites lose their uplink far more often than they lose local power: a radio path fades, a cellular cell congests, a fibre is cut. Protocols are designed for exactly this. DNP3 buffers class 1/2/3 events with quality flags and time-stamps so that, when the channel returns, the RTU replays everything the master missed in correct chronological order; the same store-and-forward behaviour exists in IEC 60870-5. Real products show how large this buffer can be - a Schneider SCADAPack x70 with firmware 9.8.6 or later can be configured for up to 100,000 buffered events across DNP3 and IEC-104, whereas the legacy E-series retained around 40,000.
Buffered events live in RAM and are only as durable as the power that retains them. If the uplink is down and the auxiliary supply then also fails, an unprotected RTU loses its unsent event history, and the operations centre can no longer reconstruct what happened on the feeder during the disturbance - precisely the interval engineers need most for fault location and protection analysis. Back-up power keeps the buffered events, the time-stamps and the real-time clock coherent through a combined communications-and-power interruption, which is the data-integrity case for a reliable secondary source rather than a mere convenience.
When the auxiliary supply crosses its under-voltage threshold, a well-designed RTU executes a last-gasp sequence in a matter of tens or hundreds of milliseconds: it detects power-fail, finishes or aborts the current scan, writes registers and event queues to non-volatile or retained memory, stamps the outage time, and - if the communication path still has energy - transmits a power-fail object or message so the master knows the site went dark rather than simply falling silent. Some designs use a small hold-up capacitor for this instant action; others rely on the battery that also provides extended back-up.
The last-gasp is a hard real-time load with a hard energy budget: a defined current for a defined maximum time, guaranteed at end of battery life and at the lowest rated cabinet temperature. It cannot be left to whatever charge happens to remain. Separating the instant last-gasp reservoir from the long-duration autonomy reserve - or proving that one battery covers both at worst case - is a core design decision, and it is where the internal resistance and pulse behaviour of the chosen chemistry decide whether the rail holds long enough for the fail telegram to leave the modem.
Beyond the last-gasp, many sites require the RTU to keep operating for hours or days without the auxiliary supply so it can record the evolution of an outage and resume reporting the instant the uplink returns. Recloser and pole-mounted controller specifications illustrate the order of magnitude: field controllers are commonly offered on 24 V DC (two 12 V blocks) or 48 V DC (four 12 V blocks) with battery options around 1.2, 7, 17 and 24 Ah, giving on the order of 48 hours of operation at a 20 W load for the larger configurations. These are published, chemistry-agnostic load examples rather than NiMH claims, and they frame the autonomy targets a secondary source must meet.
Sizing such autonomy uses the same discipline as a 24 V DC industrial UPS: average load current, required hold time, an efficiency term for the DC conversion and a Peukert-style derating (a factor near 1.25 is used for typical VRLA estimates) so that capacity is not taken at face value. The first animated figure contrasts the layered RTU current - flat scanning baseline, repeated radio bursts and the final last-gasp peak - while the second shows how buffered events accumulate during an uplink outage and drain only when the back-up source also fails.
Most RTUs sit in sealed or ventilated outdoor cabinets exposed to solar gain in summer and sub-zero mornings in winter, at sites visited by a technician perhaps once or twice a year. That reality makes the maintenance profile of the back-up chemistry a first-order selection criterion. Traditional valve-regulated lead-acid blocks are cheap and familiar but are temperature-sensitive in float service, lose capacity in the cold, dry out and require periodic replacement and resistance checks; lithium chemistries offer density but bring charging-temperature limits and more elaborate management.
Sealed NiMH occupies a useful middle position for the control cabinet: it tolerates a wide temperature span, can accept some charge below freezing where lithium cannot, contains no free liquid, uses simple charge termination, and its higher energy density than lead-acid lets a compact pack deliver last-gasp plus moderate autonomy in a crowded DIN-rail enclosure. Battery temperature monitoring - already a feature of RTU battery modules - lets the charger adapt and protects service life. NiMH is not a replacement for a multi-kilowatt-hour station battery; it is a strong fit for the distributed, modest-power, low-maintenance edge controller.
The load-profile analysis yields a layered specification: a guaranteed last-gasp energy and power budget at end-of-life and minimum temperature; retention of RAM, event buffer and RTC through combined power-and-communications outages; an autonomy target derived from measured quiescent and radio duty with a Peukert derating; low and stable internal resistance so modem transmit peaks do not collapse the rail; a charge regime compatible with continuous readiness and cabinet temperature; and a maintenance interval matched to remote-site access. The second paper turns these requirements into a concrete pack and charging architecture, comparing NiMH honestly against VRLA, supercapacitors and lithium, and the third paper maps the conformance, environmental, EMC, cybersecurity and battery-standard evidence an RTU vendor must assemble.
Treating back-up power as part of the data-acquisition chain - rather than as a generic UPS afterthought - is the mindset that separates an RTU which quietly preserves the complete event record through every disturbance from one whose master station is left guessing.
Weijiang Power designs and manufactures sealed nickel-metal hydride cells and matched industrial packs for remote, off-grid and safety-related equipment, and supports OEM partners with IEC 61951-2 performance files, IEC 62133-2 safety evidence, pulse-load characterisation, wide-temperature testing and charger/pack co-validation. Tell us your duty cycle, peak current, temperature envelope, autonomy target and the standards your product must meet, and our engineers will specify a cell-and-pack combination that protects runtime, reliability and service life. Review the range on the products page.
