Home › Package units
What is on your roof, and what to do about it
Written for Simi Valley and the valleys either side of it, where a large share of the housing has the entire heating and cooling system in one box on the roof. It covers what a package unit is, what a replacement actually involves on the day, how long one should last, and whether converting to a split system is worth the money. There is nothing to buy at the end of it.
First, which do you have?
Stand in the street and look at the roof. If there is a single large metal box up there, roughly the size of a chest freezer, you have a package unit — the compressor, both coils, the blower and the heat source are all inside it, and the ducts drop straight through the roof into your ceiling. There will be no outdoor unit at the side of the house and no furnace in a closet or the attic.
If there is a square unit with a fan on top standing on a pad at the side of the house, that is a split system: the condenser is outside, and a furnace or air handler is in a closet, the attic or the garage, joined to it by a pair of insulated copper lines.
Both are ordinary, both are serviceable, and neither is a problem. But almost every question that follows has a different answer depending on which one you own, and a surprising number of people have never been told.
Work out where you stand
Pick the situation you are actually in and the thing you are considering. Everything below is written out in full either way — the picker only narrows it.
Before you pay to convert a package unit to a split system, read this
It is the larger job, so it is the one you will most often be encouraged towards, and on most Simi Valley tract houses it is the wrong answer. Here is what it actually involves.
The indoor half has to go somewhere. That is the part people underestimate. A slab house with a low-pitched roof frequently has no closet deep enough and no attic with the clearance and the structural support for an air handler and its drain pan. Creating that space means losing a closet, building a platform, or finding out mid-job that neither works.
Everything else is new too. A condenser location and pad outside, a refrigerant line set run through the structure, a condensate route with a secondary pan and a float switch where one has never existed, new electrical to two places instead of one, a plenum transition from the air handler onto duct runs that were built to come down from above, and the roof opening where the unit used to be closed up and made good by a roofer.
What you get for it: equipment out of the sun and easier to service at ground level, a quieter house, and a wider choice of equipment. Those are genuine benefits and on some houses they justify it — particularly a two-storey house, a house where the roof is being replaced anyway, or one where there is already a usable closet.
What you do not get: a dramatically cheaper bill. The efficiency difference between a current package unit and a current split system is far smaller than the conversion cost, and anybody who tells you the conversion pays for itself in energy is doing arithmetic you should ask to see.
Our position: we will quote a conversion if you want one, and we will tell you first, in writing, whether we think it is worth it on your house. On the majority of single-storey package-unit houses in this valley, our answer is no.
Not knowing is normal, and it is the cheapest gap to close
Most people have never been told what they have, because nobody has had a reason to tell them. It matters because every quote, every price comparison and every piece of advice you read online assumes one or the other, and advice written for a split system is actively misleading if you have a package unit.
You can settle it from the street in thirty seconds — one box on the roof and no outdoor unit beside the house means a package unit. The data plate on it gives the model, the capacity and the year, and the year is the single most useful number you can know about your system.
What follows: once you know, you can judge a quote. If somebody quoting you talks about the furnace in your attic and you do not have one, you have learned something important about that quote.
Old and working is a position of strength, not a problem
A package unit in this climate typically gives twelve to eighteen years, and plenty reach twenty. If yours is working, you have time, and time is what lets you make a good decision instead of an urgent one.
The useful thing is to find out how close to the end it really is: what the pressures and amp draws say, whether the coil is clean or packed, whether the heat exchanger is sound, and whether the cabinet and the curb are still weathertight. That is a maintenance visit, not a survey, and it costs a fraction of what a rushed August changeout costs.
What we would not do: tell you to replace a working unit on age alone. The efficiency case is usually weaker than it is presented, and a unit that is genuinely finished announces itself in the readings.
Failed outright, which means the decision is now and the pressure is on
The question is repair or replace, and it turns on two things: what has failed, and how old the unit is. A contactor, a capacitor, a fan motor or a board on a ten-year-old unit is a repair and an easy call. A failed compressor or a cracked heat exchanger on a fifteen-year-old unit is the end, and spending on it is spending twice.
The middle cases are where people get hurt. A compressor on a twelve-year-old unit with a clean coil and a sound cabinet can be a reasonable repair. The same compressor on the same unit with a corroded cabinet and a curb that needs work is not. Nobody can tell you which from the driveway.
What matters right now: do not let the urgency choose the equipment. Even in August, a day spent getting a second opinion on a changeout is cheaper than fifteen years with the wrong unit on a curb that was never measured.
You can find out most of it before you own it
A general home inspection will usually report that the system operates. That is not the same as telling you what you are inheriting, and the difference is often a five-figure one in the first few years.
What is knowable before exchange: the year from the data plate, the condition of the cabinet and the coil, the state of the curb and the flashing around it, whether the roof itself is near the end — which matters enormously, because the time to replace a rooftop unit is when the roof is off — the size and condition of the return, and whether the ducts have ever been touched.
What to ask directly: how old the unit is and whether the roof has been replaced since. If both are near the end, that is a negotiating position rather than a reason to walk, and it is far better found now.
With several units, the job is sequencing, not diagnosis
Any one unit is a simple question. A roof with six of them is a different problem: they were installed at different times, they will fail at different times, and without records they all look the same until one stops.
What changes the economics is knowing the age and condition of each one separately and writing it down. Three years of readings on the same units shows you which is drifting, and drift is what turns an emergency into a line in next year's budget. Emergency changeouts in August cost more for reasons nobody controls — availability, crane scheduling, overtime — and planning around that is the single largest saving available.
What we would do: report every unit individually rather than giving the roof one verdict, and propose an order of replacement based on risk rather than on which one is easiest to reach.
Converting to a split system
Genuinely worth considering on a two-storey house, on a house where the roof is being replaced anyway, where there is already a usable closet or a proper attic, or where the roof unit is in a position that makes service dangerous or disruptive.
- What it costs: substantially more than a like-for-like changeout, and the gap is mostly labour and builder's work, not equipment.
- What decides feasibility: whether there is anywhere to put the indoor unit. Everything else is solvable; that one is not always.
- What it is not: an energy-saving measure that pays for itself. Ask to see the arithmetic from anybody who says otherwise.
- Who else is involved: a roofer to close the opening, and possibly an electrician. Those should be named in the quote, not buried in it.
Replacing it with another package unit
The straightforward path, and usually the right one on a single-storey house. It is not quite a straight swap, and the differences are where quotes diverge.
- The curb. Current units rarely match the footprint of a unit from the eighties or nineties, so an adapter is normally needed. It should be in the quote by name.
- Sizing. The load should be calculated from the house, not copied off the old nameplate. Many of these houses have been re-roofed, re-glazed or insulated since, and the right unit is often smaller.
- The duct drops. Older than the unit, and frequently the actual cause of the complaint. Worth opening and looking at while the unit is off.
- Making good. Flashing and roofing around the curb is a roofer's work. A quote that is silent about it is a quote with a gap in it.
Access, cranes and what happens on the day
This is the part that surprises people, and it is worth knowing before a date is agreed.
- Getting it up and down. Some units can be walked up a ladder frame; many cannot, and then a crane comes. That is a scheduled, priced item with its own constraints, not a detail.
- Where the crane stands. Street width, parked cars, overhead lines and trees all decide feasibility. On some streets a permit is needed for the day.
- Roof covering. Tile is walked on carefully and some breakage is normal and should be expected and allowed for. Composition is easier. Either way the roof is disturbed and a roofer makes it good.
- How long. A straightforward residential changeout is usually one day. Finding a curb mismatch or rotten decking underneath makes it two, and an honest contractor will have warned you that this is possible.
How long it should last, and what shortens it
Twelve to eighteen years is the usual range in this valley, with plenty going longer. What moves a unit to the short end of that range is almost entirely within your control.
- A starved return. The single biggest killer here. An undersized return or a filter denser than the system was built for makes the blower and the compressor work permanently harder.
- A packed condenser coil. Full sun, dust and brush debris. A coil that cannot reject heat runs the compressor hot every afternoon all summer.
- Blocked condensate. On a roof unit this shows up as a ceiling stain long after the damage started.
- Nobody ever going up there. Out of sight genuinely is out of mind, and a once-a-year look costs far less than the years it adds.
Annual service is not an upsell on this kind of equipment. It is the difference between twelve years and eighteen.
Tell you what you have, for nothing, on the phone
Describe what you can see from the street and we can almost always settle it in one call. If you can read the data plate, the model number and the year tell us the rest.
Beyond that, if you want it looked at properly, a first maintenance visit establishes the age, the condition, the curb, the return size and a set of baseline readings, and leaves you with all of it written down. From then on you are judging quotes against facts rather than against a salesperson's description.
Establish how much life is actually left, and then leave it alone
One thorough visit: pressures, superheat and subcooling, amp draw against the data plate, coil condition, heat exchanger, cabinet and curb, condensate route, and static pressure at the return. Those readings tell you whether you are looking at two more years or eight.
Then nothing, unless something changes. We would rather you replaced it in March of the right year than in August of the wrong one, and the only way to pick that year is to have the numbers.
Diagnose it properly and give you both prices
Find what has actually failed rather than what is easiest to blame, and then set out the repair cost and the replacement cost side by side with the age of the unit next to them. If the repair is sensible we will say so, and if it is money thrown at a unit that will fail again next season we will say that instead.
If it is a replacement, the curb gets measured on that visit and the access question gets answered on that visit, so the quote is a price rather than an estimate that moves.
Look at it before you exchange, not after
A pre-purchase look at the mechanical system specifically: age from the data plate, cabinet and coil condition, the curb and its flashing, the state of the roof covering around it, the return size, duct condition where it is visible, and performance readings if the system can be run.
What you get is a plain statement of what is there, what it has left, and roughly what it will need and when. If the roof and the unit are both near the end, that is worth knowing before you own both.
Inventory the roof, then propose an order
Every unit identified by position, with its age, condition, readings, curb type and access noted separately. That document is the thing that makes the next five years plannable, and it is yours — if you change contractor it goes with you.
Then a proposed replacement order based on risk: which failure would close you, which unit is drifting fastest, which can be grouped to share a crane visit. You may overrule all of it, but it should at least be written down and argued rather than decided by whichever one breaks first.
Two things worth knowing that nobody volunteers
If your roof is near the end, replace the unit at the same time. The unit has to come off for a re-roof anyway. Doing both in one visit saves a crane, saves disturbing the roof twice and saves flashing the curb twice. It is the single largest avoidable cost on this kind of system, and it is missed constantly because the roofer and the HVAC contractor never speak to each other.
A changeout in California needs a permit, and that is to your advantage, not against you. It is the only independent check that the work was done properly, and it is a matter for the City of Simi Valley's building department. A quote that proposes to skip it is telling you something about the rest of the job.
Not sure what you have? Call and describe it
Thirty seconds on the phone usually settles it, and it costs nothing. If what you need is a roofer or an electrician rather than us, we will say so.
Mon–Fri 7am–5pm