The beginner build: cold frame, low tunnel and a seed-starting bench

The most common mistake in a first greenhouse year is buying a greenhouse. A structure is the expensive, permanent, hardest-to-reverse part of growing under cover, and it is not the part that produces the result most beginners actually want. What people want in year one is transplants they grew themselves and a harvest that starts earlier and ends later. Both of those come from a propagation bench and a few square feet of covered ground, and neither needs a building.

This build is that setup, complete, at a computed total drawn from the item list below. It splits into two halves that do genuinely different jobs. Outdoors, a rigid cold frame and a fiberglass hoop tunnel put about ten square feet under glazing and one whole raised bed under fabric. Indoors, a four-shelf rack with T8 strip lights and a thermostatically controlled heat mat becomes the bench that starts every seed in the garden.

The centerpiece of this build

Outsunny

Outsunny Wooden Cold Frame, 39 x 26 in

A rigid timber frame with a hinged polycarbonate lid, 39 by 26 inches, holding two 1020 flats or a solid block of overwintered greens. It is the anchor of this build because a rigid frame is the piece that survives wind, holds its shape under snow and can be propped open by hand on a mild day. Zip-cover shelf units at half the price blow over, and the ones that do not blow over tear at the zip within two seasons.

Check price on Amazon $63.53 at the time of writing

If that is unavailable

Juwel Year-Round Cold Frame ($200.31). A larger twin-wall frame with adjustable lid stays and a proper aluminum section, worth the step up if the frame will live in an exposed position or you want to overwinter rather than only start seed.

What this build actually gets you

Season extension is measured in weeks at each end of the year, and the numbers are more modest and more reliable than most marketing suggests. A single layer of row cover over hoops holds roughly 2 to 4 degrees F above ambient on a still night. A heavier 1.2 oz fabric like the AlpineReach cover in this build holds closer to 6 to 8 degrees F. A closed cold frame with a rigid lid, which traps a genuine air volume rather than draping fabric on leaves, holds 7 to 10 degrees F on a clear night and considerably more on a cloudy one.

Translated into calendar terms in a typical temperate garden, that is about six weeks earlier in spring and six weeks later in fall for hardy crops, and it converts a hard stop into a slow decline for greens that would otherwise be finished. Kale, spinach, chard, mache, arugula and most Asian brassicas will hold in a cold frame through a considerable part of winter in zone 6 and warmer, not growing much but staying harvestable, which is a different and more valuable thing than growing.

The indoor half of the build does something separate and, in dollar terms, larger. A flat of 72 transplants costs between 30 and 60 dollars at a garden center. Two flats a year covers the entire cost of the light rack in the first season, and the rack then runs for a decade. It also removes the constraint that shapes most home gardens, which is that you grow whatever varieties the local nursery decided to stock.

The build list

Every item below is linked to its listing, and the total is computed from the same catalogue that prices the item rows, so the figure at the bottom of the table and the figure in the heading of this page cannot drift apart. Prices were recorded at the time of writing and will have moved.

Why a cold frame and a tunnel rather than one structure

They fail differently, and that is the point. A cold frame is rigid, small and heat-retentive. It is the right container for a dense block of high-value greens, for hardening off transplants over the two weeks before they go out, and for overwintering things that need protection from wet more than from cold. It is a poor container for anything tall.

A low tunnel over an existing bed is the opposite: cheap per square foot, tall enough for a crop that grows upward, and effectively disposable. Fiberglass hoops over a four by eight foot bed put 32 square feet under cover for less than the cost of a flat of seedlings. It insulates badly, it needs the cover clipped down every time wind is forecast, and it will not hold a frost-free minimum. What it does is buy the shoulder weeks on a whole bed, which the frame cannot do at any price.

Running both means the frame handles the crops where temperature stability matters and the tunnel handles the crops where covered area matters. Buying one larger structure instead gets you neither property well. The full comparison, including where a tunnel genuinely beats a hard-glazed house, is in hoop house versus greenhouse.

The seed-starting bench, which is where the money is

If you strip this build down to one thing worth keeping, it is the propagation bench. Nothing else here changes what you can grow as fundamentally, and nothing else pays for itself as fast.

Three things make a bench work, and only one of them is the light. The first is root-zone temperature. Seed germination is governed by soil temperature, not air temperature, and the difference between 65 F and 78 F soil is the difference between peppers taking 21 days to emerge and taking 8. The heat mat with an inline thermostat probe is in this build rather than a bare mat specifically because an unregulated mat sits at roughly 10 to 20 degrees F above ambient with no ceiling, which cooks a tray of lettuce as readily as it germinates a tray of peppers. The full temperature table by crop is in the germination soil temperature chart.

The second is light proximity. Seedlings stretch toward insufficient light and the resulting leggy transplant never fully recovers. The rule that matters is not wattage, it is distance: a T8 or T5 fluorescent strip belongs 2 to 4 inches above the canopy and needs raising every few days. That is why this build pairs the strip lights with a wire shelving unit rather than a fixed stand. Wire shelf uprights let you rehang a fixture at any height with cable ties, which is the cheapest adjustable-height rig available.

The third is the photoperiod, which is what the mechanical timer handles. Seedlings want 14 to 16 hours, consistently, and consistency is the part humans fail at. A mechanical timer costs under fifteen dollars, has no firmware and does not forget after a power cut.

What this build deliberately leaves out

A build list is defined as much by its omissions as its contents. Four things that appear on most beginner greenhouse lists are absent here on purpose.

• Skip A powered fan. At this size, opening the lid is the ventilation system, and a cold frame that overheats is telling you to prop it wider, not to buy a fan.
• Skip A heater. An unheated frame in USDA zone 6 will hold hardy greens through winter without any heat at all, and heating an uninsulated frame is throwing money at the sky.
• Skip Landscape staples in bulk. A 200 count box costs more than the hoops. Buy a small pack or use tent pegs you already own.
• Skip A pH meter. Bagged seed-starting mix arrives at a sensible pH. Test soil when you plant into ground, not into a tray.

What to upgrade first

Assume this setup works and you want to spend more next season. The order matters, because the obvious upgrade is usually not the one that changes the most.

1. 1 Light, always. Doubling the strips over each shelf costs less than any other item here and produces visibly stockier transplants.
2. 2 A thermostat on the heat mat if you did not buy the combo. An unregulated mat can push media past the temperature at which lettuce seed goes dormant.
3. 3 A second cold frame before a bigger one. Two frames let you run two different temperature regimes, which one large frame cannot.

Past those, the honest next step is not an upgrade to this build at all, it is the heated polycarbonate greenhouse build, which is a different structure with a different capability: a frost-free minimum you can rely on, standing headroom, and watering that runs while you are away. This build extends a season. That one changes which season you are growing in.

Siting the frame and the tunnel

Placement does more for an unheated structure than anything you can buy for it, because everything an unheated structure achieves comes from the sun. Three considerations, in order of how much they matter.

Winter sun angle first. The sun sits far lower in winter than intuition suggests, and a fence, hedge or building that casts no shadow at all in summer can shade a cold frame for the entire useful part of a winter day. Look at where the shadows fall at midday in the coldest part of the year, not at where they fall when you are making the decision. A frame against a south-facing wall gets both the direct sun and the re-radiated heat the wall gives back after dark, which is a genuinely significant effect on a masonry wall.

Drainage second. A cold frame with its base sitting in a puddle loses far more heat than one on free-draining ground, because wet soil conducts heat away roughly twenty times faster than dry soil does, and because the crops most worth overwintering are killed by wet at the crown more often than by cold. Set the frame on a slight rise, or on a bed of gravel, rather than in the lowest corner of the garden.

Access third, and it is not trivial. A frame you have to walk across wet grass to reach in the dark is a frame that does not get vented on the warm day and does not get closed on the cold night. Both of those are the actual failure modes. Put it where you will walk past it. There is more on this in the siting and orientation guide.

A first year, in order

The equipment is only half of it. The sequence below is what the build is designed to support across one full year, and it is the sequence a first-time grower most often gets wrong by starting too early rather than too late.

Late winter, indoors: the light rack starts. Onions and leeks first, then peppers and eggplant, which are slow and need the head start, then tomatoes about six weeks before the last frost date, and finally the fast brassicas and cucurbits which resent being started early and sitting in a pot. Working backwards from your own frost date is the whole trick, and the frost date planting calculator does that arithmetic for each crop.

Early spring, outdoors: the cold frame takes its first job, which is hardening off. Transplants moved straight from a light rack to open ground are checked badly by wind and UV. Two weeks in the frame with the lid progressively propped further open is the whole process, and it is covered in detail in hardening off seedlings.

Late spring through summer: the frame is empty and the tunnel hoops are stored. This is the correct state. Do not leave a cold frame closed and empty in summer, because the temperatures inside will degrade the polycarbonate faster than anything else it experiences.

Late summer: this is the moment the whole build turns on, and the moment almost everyone misses. Winter greens must be sown while there is still enough day length and warmth to grow them to a harvestable size. They will not grow to size in the cold. Sow the overwintering crops in late summer, get them established, and the frame and tunnel then hold them rather than grow them. Missing this window by three weeks is the difference between harvesting through winter and looking at seedlings that never sized up. Winter greens in an unheated greenhouse covers the timing crop by crop.

Running costs, honestly

The outdoor half of this build has no running cost at all. No power, no fuel, no water beyond what the bed needs anyway. Replacement row cover every three to four seasons is the only consumable, and a torn cover is usually the result of removing it in wind rather than of age.

The indoor half uses electricity, and considerably less than people expect. Six 2 ft T8 strips draw roughly 60 to 70 watts in total. At 16 hours a day that is about 1.1 kWh per day, or roughly 33 kWh over a 30 day propagation month. At typical residential rates that is somewhere between four and six dollars a month, running for the two or three months of the year the rack is in use. The heat mat adds roughly 20 watts, and only intermittently once its thermostat is satisfied.

Set against that, the seed itself is a few dollars a packet and produces far more transplants than any household needs. The economics of home propagation are not close, which is why the light rack rather than the structure is the part of this build worth buying first if you cannot buy all of it at once.

Working out the figures for your own house and season? The Greenhouse Build & Growing Planner is the paid version of these pages: 8 printable worksheets you fill in with your own numbers, plus the full PDF, $29.