Last week I put Otto’s head on the block (which has an utterly different connotation here than in, say, the French Revolution) and torqued it down. I thought that, given one uninterrupted evening, I should be able to button everything up and get Otto running.

As Woody Allen said, ah, if life were only like that.

History will record that one of the legendary events in the 21st century vintage-BMW world occurred when the head gasket on Paul Wegweiser’s 1972 2002tii (the F Bomb) blew at the Vintage in 2012, and Paul and Ben Tongsai replaced it in the parking lot of the hotel in under two hours. I was there, right in the middle of it all, performing an absolutely central three-pronged role: holding a flashlight, twisting paper towels and handing them to Paul so that he could clean out the threads in the head bolt holes, and finding beer for the actual mechanics once they were done.

Joking aside, it was the sort of event that forever skewed my perspective of what is possible—or even reasonable.

If I can go full-left-brain on you for a moment, the goal of replacing a head gasket at night in a parking lot is very clear: to replace the head gasket in the parking lot and be able to drive the car out of the parking lot. There will be no mission creep. There will be no “while you’re in there.” You pull the head off with both manifolds still on it, clean the block and head surfaces as best you can, slap on the new head gasket, re-use everything else, drop the head and manifolds back on, torque it down, reassemble, fill, time, done.

With Otto, as you know, what began as “the head gasket must be bad, just slap a new one in there” turned into “the head is both warped and cracked,” resulting in my locating a replacement head and having the machine shop tank it, mill it, and give it a fresh valve job. Then, prior to reassembly, I found that the rockers and shafts were pretty badly worn, so I replaced them with new parts. The assembled head was downright purty.

Now, then: With an eat-off-it-clean, newly rebuilt head, do you really think that I was going to slap on an intake manifold that had a 42-year buildup of carbon in it? I’m a hack, but this would’ve been over the line even for me.

I have an inexpensive parts-washer on wheels, one of those $120 jobs from Tractor Supply; I have it filled with kerosene. I spent two evenings standing in front of it, cleaning the lower and upper sections of the intake manifold, the four curved metal intake plenums, the injection throttle body, and some other ancillary pieces, with a variety of metal and nylon-bristled brushes. This wasn’t to make anything pretty and shiny; it was simply to keep abrasive crud that had been accumulating since Nixon was in office out of a squeaky-clean head.

Once I got the intake-manifold parts cleaned of carbon, I needed to scrape the old intake gaskets off the faces of the plenums so that I could replace them with fresh ones. Usually I can just work a single-edged razor blade good and close to the metal surface and get most of the gasket off in one pass, then scratch the remains off in small sections with the blade, but these things had bonded to the plenum surfaces like they were put on there with epoxy. I started using the little Scotch-Brite wheel on the Dremel tool to clean the surfaces. This worked well, but the wear rate of the little wheel was very high, with one wheel pretty much used up after doing both surfaces of one intake plenum. And I had three more plenums do.

At this point I looked at the shredded gasket material and had a horrible thought: It’s almost certainly original, so it’s probably asbestos. Damn!

Although very few of these cars are still running on their original brake or clutch linings, I try to be careful about this stuff, and this caught me by surprise. I’d just raised a cloud of dust. Damn! I donned a mask and a Tyvek suit for the remainder of the work, and switched from using the little wheels on the Dremel to a full-size nasty-looking heavy-duty Scotch-Brite wheel on a drill, the kind of thing you’d use to take off loose, rusty scale. Even using that, it was gruesomely slow work.

Finally, however, with the pieces all nice and clean, I could begin the assembly.

The lower part of the intake manifold bolts to the studs on the head; the upper part stands off vertically from the lower part on a few brackets. Then you lay the plastic injection lines in place and attach them to the injectors. The linkage and the throttle body can be reattached, as some of the attachment points are easier to reach with the intake plenums out. Once these things are done, the intake plenums slide onto the studs on both the lower and upper parts of the manifold.

But plenum #1 was misaligned. I couldn’t get it to slide onto the studs without smacking it with a rubber mallet. Then, once it was on, I couldn’t get it to seal against both the top and bottom gaskets at the same time. It didn’t make sense; was something cocked? I took all the plenums off, loosened the brackets that hold on the upper part of the manifold so that it was free to move, test-fit the plenums, got #1 to slide onto the studs much easier, verified that it was sealing, tightened all the plenums down, and then tightened the upper manifold-bracket bolts, which were now difficult but not impossible to reach.

It was only afterward that I saw the source of the problem: tii’s don’t have a mechanical fuel pump, so they don’t need a hole on the intake side of the head for the pump pushrod. On an original tii head, this hole isn’t there—but if you’re using a replacement head, it most likely has the pushrod hole, so you need to block it off. The block-off plate I was using, it turned out, was hitting part of the upper intake manifold. If and when I have cause to pull things back apart, I’ll sand down the corner of the plate where it’s hitting, but for now, since I’d gotten things to seal, I just left it.

When decapitating an engine, some amount of crud is certain to fall down into the oil, timing-chain, and coolant passageways. It’s best to do whatever you can to wash it out. I take the oil drain plug out of the pan, take a couple of quarts of lightweight oil like 5W30, and pour it into the head with the valve cover off—straight down the passageways on the exhaust side, and straight down over the timing chain—to try to wash out stray particulate matter. It’s not nearly as good as hot-tanking a bare block, but hey, it’s much better than nothing. With the oil flushed, I re-inserted the drain plug, filled the engine with fresh 20W50, pre-filled the oil filter, and screwed it to the side of the block.

To clean stuff out of the coolant passages, I removed the coolant drain plug on the exhaust side of the block, connected all the coolant hoses, stuck a garden hose in the radiator, and flushed. I then drained the radiator, replaced both drain plugs, and refilled with coolant mixed with distilled water.

With oil and coolant in the engine, the next precious bodily fluid to deal with was gas. Tii’s have a fairly primitive high-pressure fuel system, with many opportunities for leakage, particularly since I’d just had the fuel line disconnected and the injection lines off the car. My routine is to grab a fire extinguisher, pull out the fuel pump fuse, crack the key to ignition, then touch the fuse to its holder. In this way, you can instantly pull the fuse back out to stop the fuel pump if you see fuel leaking.

And see fuel leaking I did—all over the top of the injection pump. The engine wasn’t running yet, so fuel couldn’t be coming from the injection lines. It had to be coming, I thought, either from the check valve at the back of the injection pump or the return line connected to it. This made sense, as I’d had both of those off to get clearance for the recalcitrant #1 plenum.

But when I wiped the gas off the top of the pump, touched the fuse to its holder, and looked again, gee, it sure looked like gas was coming from the top of the pump, not just creeping there! I did this several times before I localized the leak; it seemed to be coming from the rubber O-ring under the cap for the #3 suction valve, which was weird, because I hadn’t touched any of these valves.

Fortunately, I stock these O-rings in the garage. I took a 6-mm Allen socket, pulled the cap, took off the old O-ring, installed a new one, and tightened the cap back down.

I tried again: The pump was still leaking from the top.

It turned out that three of the four O-rings had somehow spontaneously chosen to go south during Otto’s decapitation. I changed all four; the leak stopped. It was weird, but, hey, old cars, right? It’s sort of like what my childhood pediatrician used to say about me. My mother would say, “He’s never done this before.” The pediatrician would say, “He’s never been this old before.”

The last thing to do before attempting fire-up was verify the presence of oil pressure. You might think that this is trivial, and really, it should be—just disconnect the fuel pump and the ignition coil and crank the engine until the oil-pressure light goes out. If you leave the spark plugs out, the engine cranks faster, since it’s not pushing against compression.

In practice, however, engines that have been apart sometimes take a surprisingly long amount of time to build oil pressure. Otto has both the stock oil-pressure light as well as a pressure gauge, but I cranked and cranked the engine, and never got the gauge to budge or the light to go out.

Although I’ve had this happen before, I did some reading on bmw2002faq.com to get reassurance. The guidance was that, with the engine running, it may take as long as twenty seconds for oil to travel all the way up to the pressure sensor at the back of the head, and for all the air to be purged from the oil passageways so that pressure hits the sensor.

I put the spark plugs in, held my breath, turned the key, and Otto instantly fired into existence—but this event was overshadowed by my watching the oil-pressure gauge like a hawk.

After twenty seconds, it still hadn’t budged, so I shut the engine off.

Things can go wrong and cause zero oil pressure: The oil filter itself may be bad. The pressure-relief valve in the oil pump can be stuck. The splines on the gear on the front of the pump can strip, making the gear spin while not actually turning the pump. These possibilities are to be taken seriously.

The first thing to check is whether the sensor wiring is okay. The sensor simply makes or breaks the warning light’s connection to ground; when there’s no oil pressure, the circuit is complete and the light goes on. As pressure builds, it forces contacts in the sensor apart, breaking the connection to ground and causing the light to go out. So, with the ignition on, disconnecting the wire of the oil-pressure sensor should make the light go out, and grounding it should make it go on. If it doesn’t go out when disconnected, there’s likely a short to ground somewhere. I did this, and the wiring tested out fine. Note that this checks the wiring, but not the sensor itself, and certainly not the pump.

An easy test of oil-pump function is to pull off the oil filter, put a catch basin beneath the oil pan, disconnect the ignition and fuel pump, turn the key to spin the engine, and see if oil spits out where the filter was. In Otto’s case, oil instantly ran down the side of the block, so the pump was working.

The next step is to unscrew the oil-pressure sensor from the head and do the same test—disable the engine from starting and crank it. Be aware that when oil comes out of the sensor hole, a lot of it comes out—and because it’s up high, it runs down the side of engine rather than being caught neatly in a pan.

And that’s exactly what happened. All of that preliminary cranking had apparently done a great job of priming the system, because in about two seconds, oil gushed out the sensor hole. On a 2002, the sensor is at the back of the head, so of course the oil ran right down onto the flange between the exhaust manifold and the headpipe. From the twenty seconds I’d run the engine, the flange had already gotten hot, and when the oil hit it, it began to smoke.

Maire Anne saw the smoke coming out of the garage.

I had to reassure her that the place wasn’t actually on fire. With the fire extinguisher at the ready, I grabbed paper towels and wiped as much off the oil as I could. I waited about ten minutes for the exhaust to cool, then sprayed the oily area down with brake-cleaner.

I then restarted the —carefully.

The oil-pressure gauge immediately sprang to attention, and the light immediately went out.

(Note that if all that oil gushed out the sensor hole but the light didn’t go out, you’d conclude that the sensor was bad). Keeping the fire extinguisher handy, I let the car warm up to operating temperature as the residual oil got cooked off the exhaust.

So: Otto is alive, but still on the table. Now that he’s had one warm-up cycle, I’ll angle-torque his head another 25 degrees, per the head-gasket instructions, re-adjust the valves, and set the timing and idle mixture. Maybe the Frankenstein analogy isn’t the best one. Maybe it’s Star Trek III: The Search for Spock, where mind and body have been reunited—but our hero doesn’t quite remember who he is.

(Next week, hopefully Otto is discharged.)—Rob Siegel

Rob’s book Memoirs of a Hack Mechanic is available through Bentley PublishersAmazon, and Bavarian Autosport—or you can get a personally inscribed copy through Rob’s website: www.robsiegel.com. His new book, The Hack Mechanic Guide to European Automotive Electrical Systems, can be pre-ordered from Bentley Publishers. Use the coupon code “BMWCCAELECTRIC” for 30% off list.