Repair or Replace? How to Decide

Not what it fixes. What it buys, measured in years of service.

This reframing does most of the work, because it makes two very different repairs comparable. A set of springs on a sound door buys a full cycle rating — and you get to choose that rating, which means you're partly buying the interval rather than just accepting it. A patched section on a door whose other sections are equally tired buys until the next one goes, which nobody can predict. Same category of expenditure, wildly different products.

The follow-up question is what the repair depends on. A repair whose life depends on something you haven't checked is not really a known quantity. New springs depend on the door not getting heavier. New rollers depend on the track being true. A new section depends on the sections either side of it being sound enough to hinge to. Ask what each one leans on and you'll quickly see which repairs are solid and which are optimistic.

How you can tell: If nobody can tell you roughly what the repair buys, the diagnosis isn't finished. That's not an unreasonable question and any straight answer will be given in intervals rather than in reassurance.

This is the question that determines whether spending now is wasted if you replace later, and it's the one people find genuinely surprising.

What generally carries over: the opener. It bolts to a new door's top section and gets its travel and force reset, and a healthy opener runs a new door perfectly well. Remotes, keypads and wall controls come with it. The jamb framing usually stays. The slab obviously stays.

What generally doesn't: the springs, because they're wound and sized to a specific door weight and a different door needs a different set. The cables and drums, which are matched to the door's height and the spring system. The track, which is sized and shaped for the door's section height and is normally replaced with a new door rather than reused — reusing thirty-year-old track on a new door is a false economy that shows up as noise within a year. The rollers, hinges and brackets, all of which come with the new door. And the bottom seal.

So the practical rule is this: money spent on the opener transfers. Money spent on the counterbalance and the running gear does not. That doesn't make spending it wrong — a broken spring has to be dealt with either way — but it tells you where the sequencing question actually bites.

How you can tell: Look at what you're being asked to spend on and sort it into two piles: things bolted to the door, and things bolted to the building or independent of the door. The first pile leaves with the old door.

Usually they aren't, and this is where most of the confusion in this decision lives.

Consider two doors. The first is twelve years old, opens ten times a day, and has just broken its second spring. The second is thirty-two years old, opens twice a day, and is still on its original springs. Read the years and the first door looks healthy and the second looks finished. Read the cycles and it's the reverse — the first has done more work in twelve years than the second has in thirty-two.

Neither of those tells you about the sections, which is the third and separate thing. The twelve-year-old door has done a lot of cycles but its steel is young. The thirty-two-year-old door has barely worked but has had three decades of heat and ultraviolet exposure on its finish and thirty-two years of daily expansion and contraction on its fasteners.

What you actually want is three readings: the cycle load the door sees, the condition of the counterbalance, and the condition of the sections. They're independent. Get all three and the decision is straightforward. Get one and you'll draw the wrong conclusion from it about half the time.

How you can tell: Ask your technician for those three separately rather than for a verdict. 'What's the spring situation, what's the section situation, and roughly how many cycles is this door doing a year' is a better set of questions than 'is it worth fixing.'

We don't publish figures, and on this page in particular that's not evasiveness — a number quoted without knowing your door would be worse than no number. What's genuinely useful is knowing the shape of each side, so that when a quote arrives you can see what's driving it and ask about the right things.

On the repair side, the variables are: which component, how many of them, whether the counterbalance has to be handled to get at it, whether the part is current or has to be sourced, and whether anything has to be corrected first — track that's out of plumb, a jamb that no longer holds a lag screw — before new parts will last.

On the replacement side, the variables are: single or double door, construction and steel gauge, insulation level, stock design versus custom, glazing, and what has to be done to the opening. The counterbalance comes with it, sized to the new door, and so does the entire running gear.

The two lists overlap less than people assume, which is exactly why the percentage rule fails. They're not two prices for the same thing. They're two different products with different lifespans, and comparing them without knowing what each buys is comparing the wrong numbers.

How you can tell: When you get a quote, ask which line is the biggest and why. It's almost never where people expect, and the answer usually explains more about the decision than the total does.

The specific failure mode this page exists to prevent. It goes like this: something breaks, the smallest possible fix gets done because a decision feels premature, and eighteen months later the door is replaced anyway. The money spent in between bought nothing, and it was spent under time pressure.

The trap isn't cheap repairs. Cheap repairs are frequently correct. The trap is making a small decision to avoid making a large one, when the large one is already available to make.

The way out is sequencing. If there's a real possibility you'll replace the door within a couple of years, that possibility should be resolved before you spend on the counterbalance and the running gear, not after — because those are the components that don't carry over. Conversely, if you're confident you're keeping the door, then spending properly on it is the right move and half-measures are the expensive path.

The one thing that's always worth doing, either way: the safety-relevant repairs. A frayed cable, a broken spring, a bottom bracket that has moved. Those get dealt with regardless, because the alternative isn't saving money, it's a door that fails while somebody is underneath it.

How you can tell: If you catch yourself choosing the minimum option specifically because you can't face the bigger question, that's the trap. It's a very common and entirely human place to be, and naming it usually breaks it.

There's a second set of reasons that has nothing to do with repair economics, and pretending otherwise is why some of these decisions never resolve.

Insulation is the big one here. A single-layer door is the largest uninsulated surface in most Valley garages, and if you've started using that space as a workshop, a gym or an office, no repair on any list addresses it. Noise is another — some doors are loud because of their hardware, which is fixable, and some are loud because of how they're built, which isn't. Security, where an older door with a compromised lock or an easily reached emergency release is a genuine consideration rather than a sales point. And appearance, which people apologize for raising and shouldn't, since the door is usually the largest single element on the front of the house.

Keep these in their own column. A door with three years of structural life left and a repair worth doing is still a door with three years of life left, whatever else you'd like it to be. But if the second column is long and everything on it points the same way, that's real information and it belongs in the decision.

How you can tell: The test from the triage section: would you still want to replace it if it were working perfectly? If yes, you're not making a repair decision and the mechanical arguments are beside the point.

A legitimate outcome of this method, and one we reach fairly often.

Sometimes the inputs aren't there. The door hasn't been assessed properly, the springs' condition is unknown, nobody has checked whether the sections are sound, and the decision is being made on a symptom. In that situation the correct next step is neither repair nor replace. It's finding out.

What finding out looks like: a proper balance check by hand, a read on the spring size and remaining condition, an inspection of every section's interior face, an assessment of the rollers, hinges, bearings, cables and track, and a look at how the door sits in the opening. That's an hour or so and it turns a decision made on fear into one made on information.

It's also, frequently, how people discover that the door is in considerably better shape than they assumed and that what it needs is an afternoon of attention. That happens more often than the reverse.

How you can tell: If you're weighing two options and you can't answer the triage questions at the top of this page, you're not ready to decide yet. That's not indecision, it's the correct reading of the situation.