Understanding the 1906 Event and What It Actually Means for the Bay Area Today
The 1906 San Francisco earthquake is one of those events that everybody has heard of but almost nobody really understands in terms of the mechanics behind it. I was reading through some old USGS reports last week and realized how many misconceptions still circulate about what happened, why the fire was worse than the shaking, and what this means for anyone living in the region now. The earthquake struck at 5:12 AM on April 18, 1906. The rupture started near Cape Mendocino and ran northward along the San Andreas Fault for roughly 477 kilometers before stopping near the Oregon border. That is an enormous segment. Most people don't realize the fault slipped by as much as seven meters in places. The city of San Francisco sat about 80 kilometers from the epicenter, which is close enough to feel everything and far enough that the initial P-waves gave maybe thirty seconds of warning before the destructive S-waves arrived.
The Great San Francisco Earthquake
Here is the thing that trips people up: the shaking itself was not the primary cause of deaths. Fire was. When the earthquake hit, it snapped water mains everywhere. Fire stations were damaged too. Within minutes, multiple separate blazes had started across the city. Without pressurized water, firefighters could only use what was already in their tanks and what they could pull from nearby sources like the bay. The wind from the west made it worse, pushing flames toward the center of the city. The fires burned for three days. An estimated 3,000 people died that way. The shaking killed far fewer. I spent a weekend going through archived photographs from the Reconstruction era, and what struck me is how little most people know about the aftershock sequence. There were over 600 recorded aftershocks in the weeks that followed, with several exceeding magnitude 6.0. The first major one, about magnitude 6.5, hit on April 19 and caused additional structural damage to buildings that had already been weakened. This matters because it reshaped the entire building code philosophy in California. Post-1906 construction had to account for secondary events, not just the main shock. Modern codes still carry that lesson. A counter-intuitive detail that most sources gloss over is the role of the local geology. The area that got hit hardest wasn't the bedrock hills. It was the fill material. Lots of it. San Francisco had been expanded by dumping garbage, rubble, and dredged bay sediment into the waterfront district. That loose, wet fill behaved like a liquid during liquefaction. Buildings on it sank at different rates. The Transamerica Pyramid, built later on some of that same fill zone, actually has a deeper foundation system specifically designed to bypass the unstable layer. That engineering decision comes directly from lessons learned from 1906.
How the Event Changed Engineering and Emergency Response
Before 1906, there was no seismic design code in the United States. Builders didn't consider lateral forces from ground motion because nobody had a framework for it. John R. Freeman, who was a young engineer at the time, went on to develop the first modern seismic design criteria. He introduced the concept of treating buildings as systems with natural frequencies, which meant you had to think about how the structure's period matched the ground motion period. If they aligned, you got resonance and things fell down. That is basic dynamics now, but it was radical at the time. The California Building Standards Commission, which exists today, traces part of its lineage directly to the committees formed after 1906. The Uniform Building Code came out in 1927, and the International Building Code, which California has adapted, is its modern descendant. Every time there is a significant earthquake anywhere in the state, someone revises a section of that code. The 1971 Sylmar quake changed requirements for hospital design. The 1989 Loma Prieta quake changed things for bridges and soft-story buildings. The pattern is always the same: a quake happens, a committee reviews failures, a new requirement gets written, and contractors grumble about compliance costs until the next event. Here is a practical detail that matters if you live or work in the Bay Area. Soft-story buildings, which are typically two or three story wood-frame structures with large open ground floors for parking or retail, have been the focus of mandatory retrofit programs since the 1990s. San Francisco requires them to be braced or sheared with plywood or steel frames. Oakland did too after Loma Prieta. The problem is enforcement. I worked on a property assessment once where the owner claimed the retrofit was done, but the permit history showed nothing. The inspector's notes from 2003 confirmed open parking on the ground floor with no lateral system. We had to bring in a structural engineer to do a full evaluation before we could proceed with the purchase. That cost us about two weeks and roughly four thousand dollars. Getting the actual retrofit done ran another twelve thousand. It sounds expensive until you remember that unreinforced masonry and soft-story collapse was the leading cause of fatalities in both Loma Prieta and the 1989 event.
Get the Full Details
What the Modern Risk Actually Looks Like
The USGS runs the Great ShakeOut exercises every year, and the numbers they publish for a hypothetical 1906-scale event are sobering. They estimate around 1,800 to 3,000 fatalities in a similar scenario today, with tens of thousands injured. The shaking duration would be longer because the rupture zone would be even larger, and the population density is significantly higher. More people are also living and working in older unreinforced buildings than in 1906 relative to the total housing stock. One specific nuance that surprises people: the San Andreas Fault does not move in a straight line. The 1906 rupture followed a somewhat curved path, and the stress transfer between segments is complex. After 1906, seismicologists expected the fault to break again in the San Francisco section within a few decades. It didn't. Instead, the 1989 Loma Prieta quake happened on a different fault system entirely, the Los Gatos segment of the San Andreas complex. Stress doesn't just sit in one place. It redistributes. That is why scientists talk about seismic gaps, but those gaps shift over time. The most likely next big event on the San Andreas isn't necessarily in San Francisco. It could be further south, near Parkfield, or it could be a multi-segment rupture that starts somewhere unexpected. Another overlooked point is the utility infrastructure. The 1906 fire spread because water pressure dropped immediately. Today's underground utility corridors are far more complex, and while codes have improved, the aging water main network in certain neighborhoods still creates pressure vulnerabilities. PG&E has been replacing cast iron gas lines with polyethylene for over a decade now, which reduces rupture risk significantly, but the replacement is not complete. During the 2014 South Napa earthquake, a single gas line failure shut down service to about 10,000 homes. The principle is the same: when the ground moves, the things buried in it are the first thing to fail, and you cannot fix that from the surface quickly.
Preparing for What Comes Next
If you want a realistic checklist, skip the sensational stuff and focus on the things that actually matter. Three days of water per person. At least two weeks of food for a household. A hand-crank or battery-powered radio. A way to communicate when cell towers are overloaded or down. I keep a basic go-bag in my car that includes a Nalgene bottle, a headlamp, a paper map of the Bay Area with marked evacuation routes, and a whistle. The whistle is important because if you are trapped under debris, which is a real possibility in a soft-story collapse, you cannot always yell far enough. For structural preparedness, know your building. Check whether it has been retrofitted if it is a soft-story or unreinforced masonry structure. Look at the foundation. Is it bolted to the sill plate? Is the water heater strapped? Those two items are trivial to address and they prevent the most common injuries during a quake. An unstrapped water heater can fall over and rupture gas and water lines simultaneously. I once saw that happen in a rental I was inspecting, and the tenant had no idea it was a hazard because the landlord hadn't done a single seismic upgrade. Had it shaken that night, that heater would have been the first thing to cause a problem. There is no perfect solution here. The geology is what it is. The fault is moving. The city is built on fill. None of that changes. What changes is whether you understand the actual mechanisms instead of just the mythology. The 1906 earthquake taught us that fire kills more than shaking does. It taught us that fill soil amplifies and liquefies. It taught us that aftershocks are not optional. Those lessons are still being applied, slowly, to buildings and infrastructure built over the last century. The next event will test whether that work was enough.
