Two ways to get a temperature reading from a room back to a DDC panel — and why the real trade-off isn't "which one is cheaper," it's where the cost and the failure risk end up living over the sensor's 10-15 year life.
Every BAS point needs two things: power, and a path back to a controller. A hardwired sensor gets both from the same dedicated cable — a home-run pair pulled through conduit or fished through a ceiling cavity straight to a DDC panel. A wireless sensor gets its path from a radio protocol like Zigbee, LoRaWAN, or EnOcean instead of a cable, but it still has to solve the power problem — usually with a battery, occasionally by harvesting energy from light or a temperature differential. Neither approach removes the underlying requirement; each just moves it somewhere else. Where it gets moved determines the entire cost, install time, and maintenance profile of the point.
A hardwired point front-loads its entire lifetime cost into the install: conduit, cable, labor to fish it through walls and ceilings, and a home-run pull back to the panel. Once that's done, the point needs essentially nothing for 15-20 years — no battery, no RF path to maintain, no periodic truck roll. A wireless point does the opposite: almost no install cost or disruption, but a small recurring operating cost shows up for the rest of the point's life, in the form of battery swaps and, occasionally, a technician diagnosing why one sensor in a stairwell keeps dropping off the mesh. Neither number is "the real cost" on its own — the right comparison is total cost over the sensor's service life, in this building, at this retrofit difficulty, not the sticker price of the hardware or the install labor alone.
Wireless is almost always cheaper to install, especially in a retrofit. But "cheaper overall" has to include what happens after commissioning: a large building can easily carry several hundred battery-powered points, and sending a technician to open a ceiling tile, replace a coin cell, and re-commission a sensor is real, recurring labor — multiplied by every point, every 3-7 years, for the life of the building. Add that up over 15-20 years and the labor cost can quietly exceed what the hardwired cable run would have cost once, up front. The flip side is just as common a misread: assuming hardwired is automatically more reliable simply because it's a physical wire. A well-engineered mesh with self-healing routing — like the re-route around the elevator shaft above — can keep working even after individual nodes fail or a path gets blocked, and some protocols (EnOcean is the standard example) harvest their own power from light or mechanical actuation, so there's no battery in the system at all to eventually die. The honest answer is that both topologies fail in different ways, and the right choice depends on the specific building, retrofit constraints, and how many points are involved — not a blanket rule either direction.
Explains the real trade-off between hardwired and wireless BAS field sensors — not a simple 'which is cheaper' question, but where each topology's cost and failure risk end up living over a sensor's 10-20 year service life — using a floor-plan comparison of a home-run twisted-pair install and a self-healing wireless mesh.
A hardwired sensor is powered and read over the same dedicated conductor pair — typically 18-22 AWG twisted pair, home-run from the device location to a DDC panel or a local I/O module, run through conduit in new construction or fished through wall and ceiling cavities in a retrofit. Because it draws power continuously from the panel, it never needs a battery and can report on demand at whatever scan rate the controller polls.
A wireless sensor instead uses a low-power RF protocol — Zigbee and Thread form self-healing mesh networks where each node can relay another node's traffic; LoRaWAN and EnOcean typically use longer-range star or point-to-point links straight to a gateway. Most wireless sensors are battery-powered (commonly 3-7 year coin-cell or AA life depending on report frequency), though EnOcean-class devices harvest energy from light, temperature differential, or the mechanical action of a switch press, eliminating the battery entirely. Either way, the wireless side still needs at least one wired uplink — the gateway or coordinator itself is normally hardwired back into the BAS network, so a wireless deployment isn't "no wiring," it's "wiring collapsed down to a handful of gateways instead of one run per point."
Hardwired cost concentrates entirely at install: cable, conduit or fishing labor, and terminations, especially expensive in an occupied, finished retrofit where walls and ceilings can't simply be opened up without disruption. After that, the point is essentially maintenance-free for its service life — no battery to fail, no RF path to lose. Wireless cost concentrates at the opposite end: install is fast and cheap since there's no cable path to route, but the ongoing operating cost is real — battery replacement labor multiplied across every point, over every replacement cycle, for the life of the building, plus occasional RF troubleshooting where thick concrete, elevator shafts, or dense equipment racks create dead zones or intermittent dropouts. Neither failure mode is worse in the abstract; the right choice depends on point count, retrofit difficulty, and how disruptive a battery-swap truck roll is for that specific facility.
New construction with open ceilings and accessible chases usually favors hardwired sensors for critical, high-reliability points (space temperature serving a VAV loop, discharge air temperature) where the marginal cost of pulling one more cable during rough-in is low. Occupied retrofits, historic buildings with plaster or masonry walls, temporary or pilot deployments, and any project where minimizing tenant disruption matters tend to favor wireless. Large point counts change the math further: a few dozen wireless points are an easy maintenance program, but several thousand across a campus starts to make the recurring battery-swap labor a real line item worth comparing against a one-time cabling cost. A hybrid approach — hardwired for permanent, high-density, easily accessible points and wireless for hard-to-reach or temporary points — is common in practice rather than an all-or-nothing choice.
The sensors themselves don't, but the network they join does. Zigbee, Thread, LoRaWAN, and EnOcean sensors all report back to a gateway or coordinator device, and that gateway is normally hardwired into the BAS network (Ethernet or a BACnet MS/TP trunk) so its data reaches the DDC panel or supervisory controller. A wireless deployment reduces the wiring from one run per sensor to one run per gateway, it doesn't eliminate wiring entirely.
It depends heavily on report frequency and the protocol's power profile — commonly 3 to 7 years for a coin-cell temperature or occupancy sensor reporting every few minutes, shorter for sensors that report more frequently or that also drive a display. Energy-harvesting protocols like EnOcean have no battery to track at all, since the device generates its own power from light, a temperature gradient, or the mechanical energy of an actuation.
Concrete and masonry walls, elevator shafts, dense metal equipment racks, and even large HVAC ductwork can attenuate or block RF signal between nodes. A well-designed mesh network mitigates this with self-healing routing — if a direct hop is blocked, traffic automatically reroutes through a neighboring node — but a site survey before installation is still standard practice to identify likely dead zones and place gateways or repeater nodes accordingly.
Not automatically, but it is the more common choice for points feeding a tight closed control loop where update latency and guaranteed delivery matter most, such as the space sensor driving a VAV box's primary control loop. Wireless has closed much of that gap for many use cases, but a hardwired point never has to compete for radio spectrum, never has a battery that can die mid-cycle, and reports at whatever rate the panel polls rather than a report interval chosen to conserve battery life.
Yes, and mixed deployments are common — a building might hardwire high-density, easily accessible new-construction areas while using wireless sensors for a historic wing, a temporary pilot program, or hard-to-reach points like a rooftop or a space where opening finished ceilings isn't practical. Both integrate into the same BACnet-based supervisory layer once the wireless gateway is on the network, so the choice can be made point-by-point rather than building-wide.
Try our Smart Buildings Studio
More calculators, simulators, and guides for this discipline.