Part of the house buying criteria when I purchased a house in 2022 was either an existing shop space or enough room to construct a new shop somewhere on the property. As fate would have it, the place I settled on had a small shop attached, 20 feet by 24, behind the detached garage. While it did have power, it was in the form of a single 20 amp circuit. This meant all 220V tools, like full-size table saws or band saws, were out of the question. A power upgrade was in order.
The big wrinkle in this plan was complicated by several factors. The main one being the detached nature of the shop. The existing power, one circuit to the shop and one to the garage, was buried in the ground. The conduit used, given the distance from the panel, limited the potential sub-panel to 50A, which seemed adequate to power a nice woodworking shop. The rub is that with the recent shift to electric cars, it was obvious that an electric car charger would be needed in the garage. These usually run a minimum of 30A for the Level-2 chargers, which put a bit of a strain on the 50A budget. So plans for a new conduit ran began.
This is when the architecture of the house introduced some more complications. The house is a brick exterior and the electrical panel is in the basement, attached to a concrete foundation wall. Drilling was out of the question, as the tools involved to drill through rebar reinforced concrete are rather industrial in nature. Tapping the existing meter panel would be far easier, and as luck would have it, the meter and panel on the house were already upgraded to a 325A service. This meant that the panel was rated to have a full second set of service wires installed in the lugs. The end result? This little detached garage and shop now has access to as much power as the typical American house: 200A.
Why go for this much power when all load estimates didn’t exceed 70 to 80 amps? Cost. The price difference between the panel, wire and conduit for a 100A panel and a 200A one wasn’t enough to not just go for the full size and never worry about it for a very long time. This sort of project is not one that lends itself to incremental upgrades, so buy once, cry once as they say.
After digging through the electrical catalogs, a new challenge appeared. The design for the meter panel revealed that line (from the utility) and load (to the house) wires were in the same chamber. This is different from typical 200A meter panels which usually have them in separate chambers such that pulling the meter makes it safe to work on the load side without the danger of contacting the still-live line side conductors. Given that disconnecting the meter would not de-energize the panel such that I was comfortable working on it, I deferred to an electrical contractor to install a disconnect off the meter panel. These external disconnects are now code, so while this technically was required in any case for service panels in my area, it also also provided a safer means of working with the wiring.
With the disconnect installed and ready to go, it was time to fight the service wires through the conduit. Due to the price of PVC, the de facto 3-inch conduit was too expensive. While 2-inch was an allowed size, it was at the limit of fitting the wires through, plus the run had the full 360 degress worth of bends. There were some regrets part-way through about not spending the money and it took several days of shoving but it got through in the end and landed at the panel.
Before going any further, it was time to address the floor of the shop. It was rather cracked and chipped, and there was a board being used as a control joint. From several contractors, it was apparent that the floor was too far gone; if some work had been done over the years to repair it, there may have been a chance to not have to rip it all out. Alas, it all had to go, which wasn’t great timing weather-wise, as there was a bit of a cold-snap with the highs in the low 30s for the week the work was scheduled.
And yet the structural work was still not done! Of the 12 rafters in the shop side, only 6 had ties and there wasn’t a ridge beam (or even a ridge board). Since the rafters were on 16-inch centers, this meant the ties were on 32-inch centers, which is way too wide these days, with most trusses on 24-inch centers. Luckily the rafters themselves were in relatively good shape for being built in the 50s/60s, so there wasn’t any need to replace anything.
While adding the missing ties would probably have sufficed, as with previous parts of this project, it was probably best to retrofit them and make them full trusses. To save on labor and cost, it was more than adequate to only convert the rafters that didn’t have ties. This left the existing ties to be used to hang the ceiling panels from later.
The final truss design was essentially a Howe truss built with all 2x6 boards. This was hilariously overbuilt, as my non-engineering analysis put the load capacity of the roof at over 100 pounds per square foot, well in excess of the snow loads the Pacific Northwest sees around the Puget Sound and more than capable of holding some solar panels on down the line.
The installation was not the proper way to build trusses: on the ground and lifted into place. Building on the ground removes gravity pulling down on the load path during assembly. By installing the pieces in the air, it required the weight of the truss parts to be supported so that it would actually hold the existing roof. Since it was only six trusses, it wasn’t a huge hassle.
There was one other thing that needed addressed with the roof before calling this part done. For some reason, the ridgeline of the building was not centered, it was offset a couple feet. This meant the wall on one side is lower than the other, at just under five and a half feet. Previously there was an eighteen foot long pair of shelves that were also being used to support the extended roof depth. I wasn’t a fan of the shelves as they were, but I was unsure how necessary the support itself was.
To avoid needing a couple pillars or something, I tackled this by swapping the shelves with a wall to create a small room. This provided a slightly separate storage area for wood and the opening in the wall is large enough to allow boards up to about twelve feet long to be stored. While full-size sheet goods could also be slid in, it’ll probably be less effort to just lean them up against the outside of the wall or get them broken down before storing. I also have a pretty good plywood supplier close by, so hopefully I shouldn’t need to keep much stocked on hand.
With the roof system beefed up, it was time to start leveling the ceiling so that strapping could be installed to hang the panels. Since there was almost 4 inches difference side-to-side and corner-to-corner, a pack of shims wasn’t going to cut it. After finding the high point and pulling some strings, a bunch of 2x4 blocks were attached to the ties on 16-inch centers. This provided a level set of attachment points to install some strapping, which gave more screw points for the ceiling panels for better support.
Once the strapping was set, EMT was installed on the other side of the ties as raceways for the wiring for the ceiling lights. These were going to be some 6-inch LED recessed discs that clamped themselves into their cutouts and had a short lead to a 12V driver. The drivers had 1/2 inch knockouts, making them easy to install in a line with conduit between them. The grid was going to be 5x5, to ensure there was going to be plenty of light.
Once each row of drivers was in place, next came the ceiling panels. While drywall would have sufficed for this, since the panels were going on strapping with not a lot of stud contact, there was no real possibility of hanging anything substantial from the ceiling without further reinforcement on down the line. That made OSB, basically the same price, a slightly more attractive alternative. That did make cutting the holes for the lights a bit of chore compared to just slicing through sheetrock.