How to Design Outdoor Camera Power for a Site With No Conduit Path
Every integrator eventually meets the pole. It's 400 feet from the building, it's exactly where the camera needs to be, and there is no conduit, no budget for trenching across the finished parking lot, and a customer who has already been told 'no' by two other companies. The camera is the easy part of that job. Power and backhaul are the design problem, and the difference between a system that survives five winters and one that becomes a standing service contract is decided in the survey, not the install. Here's the process I run, in order, with a worked example from the kind of remote-lot job that generates these calls.
Step 1: Distance From Nearest Network Closet
Measure first, and measure the path the cable would actually take — up walls, across ceilings, around the loading dock — not the straight line on the aerial photo. The measurement sorts the job into regimes. Under 100 m (328 ft): standard PoE over Cat6A, done, stop over-designing. 100–200 m: PoE extenders or long-reach media make copper viable. 200–600 m: this is fiber territory if any pathway exists, or long-reach alternatives if it doesn't. Beyond 600 m, or across property you don't control: wireless backhaul with local power, or solar. Also inventory what's already in the ground: legacy coax from an old analog system, an abandoned phone pair, a live 120 V circuit feeding the pole's lighting. Existing copper of almost any species is a gift — there are purpose-built ways to run both data and PoE over it, which is often the difference between a two-day job and a directional bore.
Step 2: Ethernet, Fiber, or Wireless Backhaul
Pick backhaul by distance, sightline, and what exists — in that order of preference. Fiber wins wherever a pathway exists or trenching is affordable at all: immune to lightning-induced surge on the data line, unlimited practical distance for this use, and cheap per foot once the trench exists. Legacy copper reuse is next: modern Ethernet-over-coax and Ethernet-over-2-wire gear delivers full-rate video plus PoE over cable that's been in the ground for twenty years — the NVT Phybridge CLEER line, for example, pushes PoE and data over coax to distances around 600 m (2,000 ft), six times the Ethernet spec, and their 2-wire products do the same trick over old phone pair. On the two-camera pole fed by orphaned RG-59 from a 2004 analog system, that gear turns a trenching quote into an afternoon. Wireless point-to-point is the fallback when nothing physical exists: a clean 5 GHz PtP link over a few hundred meters sustains 100+ Mbps reliably with proper antennas and clear Fresnel zone — but it solves only data. The camera still needs local power, which is why wireless always drags Step 4 into the design. Two wireless caveats from repeat offenders: survey the RF environment before committing (a lot within reach of an airport, port, or another campus's PtP links can be noisier than the aerial suggests), and mount both radios where a truck, a tree's summer foliage, or next year's inventory stack can't grow into the path — the link that commissions clean in March and dies in July usually got interrupted by something that grew or got parked.
Step 3: PoE Extender Reality
Extenders are honest tools with dishonest reputations, in both directions. The physics: each single-port extender regenerates the link and typically buys another 100 m, and cascading two or three is supported by the better product lines. What the brochure underplays: every extender in the chain is powered from the PoE budget flowing through it. Start with 802.3bt 60 W at the switch, subtract each extender's own draw (2–4 W) plus cable resistive loss per segment, and a two-extender chain to 300 m can arrive with 35–40 W deliverable — fine for a fixed dome pulling 12 W, marginal for a PTZ with a heater that wants 45–60 W in January. Do the arithmetic per device, at winter load, not nameplate. And every in-line extender is a component in a place you chose because it was hard to reach: it wants a weatherproof enclosure, a drip loop, and a note in the as-builts, because the tech who diagnoses this run in year three didn't install it. My rule of thumb: one extender is routine, two is a design decision, three means I should have found fiber or coax.
Step 4: Solar and Battery Edge Power
When no cable of any kind can reach the position, power gets generated or stored locally. Size it from the load up: a fixed bullet with IR and a wireless radio runs 10–20 W continuous; add a heater and you're at 40–70 W in winter — which is why heater-equipped cameras on solar are usually a mistake and cold-rated cameras without heaters are the right partner for off-grid positions. The sizing chain: daily load in Wh (15 W × 24 h = 360 Wh), divided by your location's worst-month sun hours (2–3 in a northern-tier December, not the 5.5 the summer survey suggested), times a 1.4–1.5 derate for panel aging, snow, and charge losses — that 15 W camera wants 250–350 W of panel, not the 100 W kit the marketplace listing suggests. Battery: 3–5 days of autonomy for stretches of overcast, so 1.1–1.8 kWh, and LiFePO4 over lead-acid anywhere that freezes — usable capacity and cycle life at low temperature aren't close. The recurring field failure: systems sized on annual-average sun that die every year in the second week of continuous overcast, get jump-started, and repeat. Winter design or no design.
Worked Example: A Remote Lot Job
The composite job: an equipment dealer wants coverage on a storage lot entrance 380 m from the building, across pavement nobody will cut. Assets on the pole position: a lighting circuit (live 120 V, switched by photocell — unusable directly for 24/7 gear without an electrician re-feeding it unswitched) and a clear line of sight to the roof. Design: two cold-rated 4K bullets (no heaters, −40°C rated) at 11 W each plus a PtP radio at 8 W — 30 W total load. Backhaul: 5 GHz point-to-point from pole to roof, engineered link budget with margin for rain fade, mounted above the light head to clear truck traffic in the Fresnel zone. Power: electrician converts the pole's photocell feed to an unswitched circuit ($600, one visit) feeding a NEMA-rated enclosure with a small UPS — 30 minutes of ride-through covers the utility blips that would otherwise reboot the radio nightly in storm season. Total cost landed under a third of the trenching quote. Had the 120 V not existed, the same design goes solar: 30 W continuous × 24 h = 720 Wh/day, worst-month 2.5 sun hours → roughly 430 W of panel and 2.5–3 kWh of LiFePO4 — still cheaper than the bore, but now a maintenance commitment.
Remote Power Compatibility Framework
| Site condition | Backhaul | Power | Watch out for |
|---|---|---|---|
| ≤100 m, any pathway | Cat6A PoE | From the switch | Nothing — don't overthink it |
| 100–300 m, pullable path | Fiber (media converter) | Local 120 V or PoE injector at head end | Converter needs its own protected power |
| Legacy coax / phone pair in ground | Ethernet-over-coax / 2-wire w/ PoE | Carried over the legacy copper | Verify pair integrity and splice condition first |
| 200–300 m, copper only, modest load | Cascaded PoE extenders | Through-budget from 802.3bt switch | Winter wattage math; enclosure at each unit |
| No path, has line of sight + local 120 V | 5 GHz PtP radio | Local circuit + small UPS | Switched lighting circuits; Fresnel clearance |
| No path, no power | PtP radio | Solar + LiFePO4, winter-sized | Heaterless cold-rated cameras only; snow on panel |
Step 5: Surge and Lightning Path
A camera on an isolated pole is a lightning rod with a lens. Every conductive line entering the enclosure gets a surge protection device rated for the service — the PoE/data line, the 120 V feed, everything — and the SPDs are only as good as the ground they dump into: an 8-foot rod at the pole bonded properly, not a self-tapper into the pole base. This is another place fiber and wireless quietly out-engineer copper: a 380 m copper run between building and pole is a induced-surge antenna that couples nearby strikes into your switch room, while fiber carries nothing and a radio link breaks the galvanic path entirely. On the jobs where I've seen a strike take out gear at both ends, there was copper between them every time. Budget real SPDs and grounding at 3–5 percent of the remote position's cost; it's the cheapest insurance on the whole design.
Step 6: Maintenance Access Plan
The position was chosen because it's hard to reach; the maintenance plan has to respect that. Spec remote reboot into the design — a managed PoE port you can cycle from a desk, or a smart relay on the 120 V feed — because the most common service action on remote gear is a power cycle, and a 40-minute drive plus a lift rental to flip a breaker is a $500 reboot. Put the radio, UPS, and any extenders on the monitoring you already run (per-port PoE draw baselines catch a failing heater or radio early), keep the enclosure serviceable from a ladder rather than a bucket truck where the mounting allows, and photograph everything inside the enclosure at commissioning so phone support can work. Solar positions add a seasonal item: panel cleaning and snow clearing on the worst-month schedule, assigned to a named person, or the winter design math you did in Step 4 gets defeated by a leaf layer.
Deployment takeaway: Survey in this order before quoting: measured cable-path distance, inventory of any legacy copper or live circuits already at the position, and line of sight to the building — those three facts pick the backhaul and power architecture for you. Then run the two winter numbers that kill these designs: deliverable watts at the camera after extender and cable losses in January, and worst-month solar harvest if you're off-grid. Monday morning: pull up your last 'no conduit' quote and check whether abandoned coax or a lighting circuit at the position would have changed the answer.
Where This Fits in a Deployment Program
No-conduit positions are where camera programs either stall or get creative, and the creative answers — legacy-copper reuse, engineered wireless, winter-sized solar — are all standard parts, not exotica, once the survey discipline is in place. Design power and backhaul as carefully as the optics and the remote positions become the most reliable views on the site instead of the flakiest. The building blocks live across our IP Cameras catalog (including cold-rated bullets from the Axis catalog that skip the heater budget), the NVT Phybridge long-reach PoE line for legacy-copper paths, plus switching and media conversion in Infrastructure and battery backup in Power & UPS. If you've got a pole, a fence line, or an outbuilding that everyone else has quoted as a trenching job, send the distances and a photo of what's already at the position — we'll help you spec the no-dig version.