WirelessHART and ISA100 beside Wi-Fi
WirelessHART and ISA100 Wireless carry slow process measurements on batteries that last years, but they share 2.4 GHz with your Wi-Fi and each network costs the other airtime, so the design job is a channel plan both can live with.
WirelessHART and ISA100 Wireless carry process measurements on battery power, a few bytes every few seconds or minutes. They share 2.4 GHz with your Wi-Fi. Where the two overlap, each costs the other airtime, and the instrument network pays in retries and battery.
What they are
Both are wireless sensor networks for process plants: refineries, chemical plants, tank farms. Neither is Wi-Fi.
WirelessHART is IEC 62591, first published in 2010 and revised in 20161. FieldComm Group describes it as wireless HART, compatible with existing HART devices, commands and tools. A network has three parts: field devices, gateways that connect them to host applications, and a network manager that configures the network, schedules communication, manages routes and monitors health. The manager can run in the gateway, the host or the controller.
ISA100 Wireless is ANSI/ISA-100.11a, published internationally as IEC 627342. Its industry body stresses native IPv6 and a tunneling feature that carries other protocols3.
Both use IEEE 802.15.4 radios in 2.4 GHz. NIST puts their signal at a nominal 5 MHz wide, against 22 MHz for an 802.11g transmission4. WirelessHART uses channels 11 to 25, and ISA100.11a adds channel 26 as an option5. The radio runs at 250 kbps6.
How they share the air
They schedule instead of contending. WirelessHART divides time into slots fixed at 10 ms, and the network manager assigns each slot to a sender and a receiver. A slot holds one packet and its acknowledgment. At 250 kbps that exchange takes about 5.1 ms, half the slot.
They hop. A WirelessHART link changes channel with every packet, and the network manager can blacklist channels where reception is poor. ISA100.11a makes the slot length configurable per superframe, offers slotted, slow and hybrid hopping, and predefines five hopping patterns.
A schedule is predictable, not fast. In a 2009 test, a WirelessHART network ran a superframe of 150 slots, so each link could transmit once every 1.5 s, and latency settled near 2 s. Its instruments reported once a minute.
What they do well
Batteries that last. One instrument maker’s WirelessHART temperature transmitter offers update rates from one second to 60 minutes7. It rates the power module at 10 years with a one-minute update rate, at 21 °C, while routing for three other devices. That is the maker’s rating at its own reference conditions.
The update rate is the budget. A column in Control notes that update rates can be as short as 0.5 s, and that battery life falls exponentially as the rate rises8. It reports one maker advising no faster than 4 s, to protect the number of devices per gateway. Every faster reading costs battery and gateway capacity.
Delivery that routes around trouble. Every WirelessHART device can route for its neighbors, which gives redundant paths in a difficult radio environment. In the 2009 test, reliability stayed at 100% over 120 hours, with packet loss below 1% for most of the run. FieldComm Group claims 30% to 60% lower installation cost than wired instruments9. That is the group’s own claim.
What they do not do
They do not carry machine I/O. The strongest objection deserves a hearing: IEC 62734 lists closed loop control among its uses, and the ISA100 body says it supports a wide range of applications, including safety. The same abstract sets the scope, though: process automation where end-to-end latencies of at least 100 ms can be tolerated, for fixed, portable and slowly moving devices.
Process loops can live with that. The Control column sizes a temperature loop whose thermowell takes 16 s or more to respond, and lands on a 4 s update rate. A PROFINET device at a 2 ms update time declares a failure after 6 ms without data10 at the default setting. The PROFINET and EtherNet/IP entries hold the machine-side numbers.
A survey of WirelessHART research puts safety and emergency functions at 10 ms, closed-loop control at 10 to 100 ms, and monitoring at 100 to 1,000 ms11. It notes the standards were developed mainly for the monitoring end. NIST adds that a mesh gains reliability at the expense of deterministic delay.
They are not a network for vehicles either. Slowly moving devices are the scope. A fleet roaming between access points is not.
Beside your Wi-Fi
The overlap is the problem. A Wi-Fi plan on channels 1, 6 and 11 leaves only WirelessHART channels 15, 20 and 25 in the gaps, and 12 of the 15 overlap Wi-Fi.
Wi-Fi costs the instruments. In the 2009 test, three access points sent nothing but beacons. At 0.8% duty cycle per channel, mean packet loss rose from 0.50% to 4.83%. At 3.9% it reached 27.20%, and mean latency rose from 2,037 ms to 2,760 ms. Reliability held at 100% through retries, and the authors note those retries cost latency, power and battery life.
NIST tested an ISA100.11a link against a Wi-Fi transmitter at 17 dBm on channel 6 with a 50% duty cycle12. At the edge of range, with the interferer 5 m from the gateway, packet errors rose from 22% to 56% and maximum latency from 2 s to 7.5 s. NIST concluded that Wi-Fi interference may degrade performance significantly, so its use has to be monitored and controlled.
The instruments cost Wi-Fi too. Low-power 802.15.4 transmissions are likely to be overlooked by Wi-Fi’s carrier sensing, so Wi-Fi can transmit on top of them. In one lab test, a Zigbee signal, which uses the same 802.15.4 radio, sat 15 dB below the Wi-Fi signal and still drove Wi-Fi 6 at MCS 7 to 100% packet errors13. See wireless cameras and sensor networks for how that looks on the floor.
What each side can do
On the instrument side, blacklist. Take the channels under your Wi-Fi out of the hopping list. Each channel you remove leaves fewer to hop across, so check how many remain.
On the Wi-Fi side, leave 2.4 GHz. The 2009 authors call moving Wi-Fi to 5 GHz the safest approach. If 2.4 GHz must stay, they suggest no more than two access points in range; channels 1 and 6 leave six clear WirelessHART channels.
Keep transmitters apart. The NIST results were worst with the Wi-Fi transmitter closest to the gateway. Do not mount a 2.4 GHz access point beside a gateway.
Put both in one inventory. NIST’s spectrum management plan holds an inventory of all known wireless networks and sources of emissions. In a process area, the hazardous-area classification also limits where your access points can go.
When Wi-Fi carries the gateway
A gateway connects the instruments to host applications over a high-speed backbone or another existing plant network. Sometimes that network is your Wi-Fi. The gateway then sits behind a fixed Wi-Fi client, a workgroup bridge.
That changes three things. Every instrument behind the gateway now depends on two networks, and a backhaul outage silences all of them at once. The bridge is a Wi-Fi transmitter next to the gateway, so run it at 5 GHz. And the readings join the airtime your Wi-Fi already shares, which the airtime guide covers. When readings stop, check the backhaul as well as the mesh: backhaul faults look like wireless faults.
When to use it
- Process measurements such as temperature, pressure, level and flow, read every few seconds to every hour.
- Instruments where cable is costly, running on battery for years.
- Monitoring, and slow process loops whose timing the control engineer has checked against the network’s worst-case latency.
Keep it away from cyclic machine I/O, safety functions and vehicles. For readings that can wait minutes, see LoRaWAN beside Wi-Fi.
What to ask a vendor
- Which standard and edition each device and gateway implements: IEC 62591:2016 or IEC 62734.
- The battery life at the update rate you need, with the reference temperature and routing load behind the figure.
- Whether the network manager blacklists channels on its own or you set the list, and how many channels remain afterward.
- The worst-case latency at your hop count and update rate, not the average.
- How many devices one gateway carries at your update rate.
- Which backhaul the gateway supports, and, if it is Wi-Fi, which band.
Give the instruments clean channels and give your Wi-Fi 5 GHz, and neither network pays for the other.
About this page
Built from 14 sources: 5 standards bodies and labs, 3 protocol owners and alliances, 4 research papers and theses, 1 vendor document and 1 other source. Researched and drafted with AI assistance. How pages are made
- First published
- Last updated
Change history (1)
- First published
Cite this page
Plain
Ben Rutter. "WirelessHART and ISA100 beside Wi-Fi." OT Wireless, published October 8, 2026. https://otwireless.com/technologies/wirelesshart-and-isa100-beside-wifi/
APA 7
Rutter, B. (2026, October 8). WirelessHART and ISA100 beside Wi-Fi. OT Wireless. https://otwireless.com/technologies/wirelesshart-and-isa100-beside-wifi/
BibTeX
@misc{rutter2026wirelesshartandisa100besidewifi,
author = {Rutter, Ben},
title = {{WirelessHART and ISA100 beside Wi-Fi}},
year = {2026},
howpublished = {\url{https://otwireless.com/technologies/wirelesshart-and-isa100-beside-wifi/}},
organization = {OT Wireless},
}