We get called out to a lot of sliding doors that do not need replacing. Somebody has decided the door is finished because it takes a shoulder to move, and nine times out of ten the door is fine and the rollers underneath it are twenty years old and full of grit.
The two-minute test that settles most of it
Before anyone measures anything, walk the door through four checks. They take less time than reading about them.
- Slide it slowly, end to end. Does it get harder at one point, or is it heavy all the way? Heavy at a point is track debris or a bent track. Heavy everywhere is rollers.
- Look through the glass toward the light, from an angle. A haze, a milky patch or visible droplets between the panes means the sealed unit has failed. This does not clean off, and it never gets better.
- Run the back of your hand around the frame on a cold windy day. Air movement at the interlock where the panels meet is normal-ish and adjustable. Air moving at the head or the jamb corners is the frame, and the frame is not adjustable.
- Look at the corners of the frame itself. Daylight at a mitre, a gap that has opened, rot in a wood sill, or a frame that has gone out of square with the opening — those are structural and they are not repairs.
Checks one and two point at repairs. Checks three and four point at a door.
Repair, or replace?
| What is happening | What it usually is | Repair or replace |
|---|---|---|
| Takes a shove to move, grinds along its length | Worn rollers, or a track packed with grit and pet hair | Repair. Clean the track, replace the rollers, reset the height |
| Catches in one spot only | A dent or a burr in the track, often from something dropped | Repair. Track cap or track repair section |
| Fog, haze or droplets between the panes | Failed insulated glass seal | Depends on the door. Some take a new glass unit; on many older panels the glass and panel are one item |
| Latch will not engage without lifting the door | Panel has dropped on worn rollers, or the strike has drifted | Repair. Adjust rollers first, then the strike |
| Draft at the head or at the frame corners | Frame seal failure or the frame is out of square in the opening | Replace. You cannot weatherstrip a frame back into square |
| Water on the floor inside after a driving rain | Blocked weep holes if you are lucky, a failed sill pan if you are not | Clear the weeps first. If it still leaks, replace |
| Condensation or frost on the inside of the frame in a hard freeze | Non-thermally-broken frame conducting cold straight through | Replace, when you are ready. It will do it every freeze |
| Rot in a wood sill or jamb | Long-term water entry, usually at the sill ends | Replace, and find out where the water came from |
| Screen sags, tears or will not roll | The screen | Repair. This has nothing to do with the door |
What the 2021 and 2022 freezes showed about old sliders
North Texas does not get many hard freezes, which is why the ones we do get are so revealing. In February 2021 the metroplex spent the better part of a week below freezing, and December 2022 dropped the same way over a couple of days. Both times, the same calls came in.
Old aluminum sliding doors — the ones in houses built here through the seventies and eighties — running with water down the inside of the frame, and in the worst cases carrying a rim of frost on the interior face of the metal. That is not a seal problem and no amount of weatherstripping fixes it. Aluminum conducts heat extremely well, so a frame with no thermal break is a direct path from outside air to inside air. When the inside surface drops below the dew point of the room, the room’s moisture condenses on it.
A modern door breaks that path — either a thermal break in the aluminum, or a vinyl or fiberglass frame that does not conduct in the first place. It is one of the clearest before-and-after differences in the whole category, and it is invisible for about 360 days a year.
The numbers worth knowing before you shop
Two figures describe a glass door’s performance, and both are printed on the label of any product sold here: U-factor, which is how readily heat passes through the whole assembly, lower being better, and SHGC, the solar heat gain coefficient, which is how much of the sun’s heat the glass lets through, lower meaning more shading.
For what counts as good enough in this region there is a genuinely local answer. The 2021 energy code as amended through the North Central Texas Council of Governments — the amendments our cities work from — sets fenestration U-factor at not more than 0.40 in climate zone 2 and not more than 0.32 in climate zone 3. The 2021 code redrew the climate-zone line right through this metroplex, so which of those two numbers applies depends on which county you are in, and your building department is the one to ask.
Two honest caveats. Those figures govern new construction and permitted work, not a like-for-like door swap — nobody is obliged to hit them on a replacement. And a U-factor describes the whole assembly, so a door that meets it on paper and is then installed out of square and foam-gunned badly will not perform like the label. Installation is not a detail here; it is most of the result.
On SHGC, the thing that matters in North Texas is orientation. A west-facing slider takes the full four-to-eight-o’clock sun through July and August, and a low-E coating chosen for solar control makes a noticeable difference to how the room behind it feels at six in the evening. A north-facing door is a much less interesting problem.
While the opening is open: sliding, or hinged?
If you are replacing anyway, it is the one moment the question is cheap to ask. Sliding wins where floor space is tight, because nothing swings into the room, and it wins on glass area for a given opening. Hinged wins on clear opening width when both leaves are open, and on the feel of the thing — a well-hung hinged door closes with a solidity a slider does not have.
For a patio that gets used constantly through the spring and fall — which is most of the year that matters here — the honest tiebreaker is usually furniture. Look at where the sofa is.
When we tell people to leave it alone
If the pitch is energy savings and nothing else, we would rather you spent it in the attic. A sliding door is a modest share of the surface area of a DFW house, and it is a much smaller hole than an under-insulated attic with leaky ducts running through it. If the door works, seals reasonably and is not fogged, replacing it as an energy measure is not where the money works hardest. Insulation almost always is.
And do not replace a door because it fogged in one panel if the frame is sound. Find out first whether that door takes a replacement glass unit. Plenty do. It is a fraction of the work and it solves the actual complaint.
The honest reasons to replace a sliding door are that it no longer keeps weather out, that the frame has gone, that it fogged and cannot be re-glazed, that it frosts on the inside every freeze, or that it does not lock properly. Those are real. “It is old” is not one of them.
The bottom line
Check the rollers before you check the price. Look through the glass at an angle before you look at brochures. Feel the frame corners on the next cold windy morning.
If those three checks come back clean and the door is just heavy, you have a repair. If the frame is the problem, replacement is the only real fix, and the thing to compare on is the U-factor, the frame construction and the quality of the installation — in that order.
Questions we get asked about this
Almost always the rollers, the track, or both. Rollers wear flat and seize after enough years, and the track fills with grit, sand and pet hair that acts like a brake. A roller set and a thorough track clean will usually make a twenty-year-old door move easily again. Replacing the door for this reason alone is rarely necessary.
The fog is a failed seal in the insulated glass unit, and it cannot be cleaned or dried out. Whether it is fixable depends on the door: many take a replacement sealed unit within the existing panel, which is a fraction of the cost of a new door, while on some older panels the glass and the panel are a single assembly. Find out which yours is before accepting a quote for a whole door.
The regional amendments to the 2021 energy code used across North Texas set fenestration U-factor at no more than 0.40 in climate zone 2 and no more than 0.32 in climate zone 3, and the metroplex is split between the two zones. Those apply to new and permitted work rather than to a straight replacement, but they are the sensible benchmark to shop against. Your building department can tell you which zone your city enforces.
If the water is on the inside face of the frame rather than the glass, the frame is conducting cold straight through — typical of older aluminum doors with no thermal break, which is exactly what showed up across the metroplex in the February 2021 and December 2022 freezes. A modern thermally broken aluminum, vinyl or fiberglass frame stops it. Weatherstripping does not.
Just the glass, where the door allows it and the frame is still sound and square. Replacing the sealed unit solves fogging and restores the insulating value without touching the opening, the trim or the wall. Once the frame is out of square, leaking at the corners or rotted, the glass is no longer the problem and a new door is the only fix that lasts.
On its own, usually not first. A patio door is a small share of the envelope of a typical DFW house compared with an attic that is under-insulated and has ductwork running through it. If the door is sound, the money does more in the attic. If the door has failed — fogged, drafty at the frame, frosting in a freeze — then replacing it is worth doing on its own merits.
