Note: Most of this article is hard fact. In a few cases, where facts were not readily available, there are researched assumptions as indicated.


In 1-2 months, Las Virgenes Municipal Water District’s Board of Directors is expected to have the *final* vote to approve the water tank: now a $9 million project (up from $6 million). If approved, and assuming the project costs don’t continue to increase, each LVMWD customer (e.g., household) — in Calabasas, Hidden Hills, Agoura Hills, and Westlake Village — is paying an average of $455 for the tank project. The chances of needing the tank pencil out to less than 1% chance (see below).

From the beginning, Las Virgenes Municipal Water District’s core reason for the water tank has been fire safety. Specifically, a triple situation happening simultaneously: hot day peak usage, major disruption of water supply, and a massive commercial/residential fire (not brush). But, the District has never answered the question of the likelihood of needing the tank. To the contrary, LVMWD’s Board President Charlie Caspary disbanded the probability analysis committee.

In terms of preparedness, the District now and always has had the option of using unfiltered water from the Westlake Reservoir, with a boil order, to handle catastrophic fires. The 3.1 billion gallon reservoir is 620 times bigger than the proposed water tank.

After talking to statisticians on how to calculate the probability properly, and determining components for each fact (or researched assumptions), the rough estimate comes out to a 0.29% chance of actually needing the tank in the next 30 years. That’s less than 1/3 of 1% chance. Note: The triple threat scenario has never happened, nor has it been even close to happening. A real probability study would have taken this rough estimate to a more detailed analysis.

All of LVMWD’s customers pay for the tank. The District has made clear that if the tank is not built, most of the money saved can be used to offset both the rate increases that LVMWD’s board has already approved or projected for the next several years. The funds could be used to offset other costs: the expected pass along rate increases from the wholesale water supplier (MWD increases are in addition to LVMWD increases), the possible near tripling of sewer rates due to the new EPA standards imposed, conservation efforts, etc…

Unless something changes or the community directs otherwise, residents and ratepayers should expect three of the five board members to vote to approve the tank project. There are no other votes on the tank after this final approval: it’s the approval to begin construction.

Westlake Revelations readers have asked about where to direct feedback. As always, voicing your opinion to elected officials is the best way. If you want to give LVMWD feedback, including support or opposition to the tank’s construction, call LVMWD at 818-251-2100 or send email to its general manager, David Pedersen As always, feel free to cc (or not) on your emails as well as others you want to inform in the community, or those you feel that can have an impact: other news outlets, or community leaders.

The Triple Threat

To be clear, this isn’t a matter of whether the community should be prepared for fire. It’s simply a matter of whether potable (filtered, ready to drink) water is used to fight a commercial fire, or if LVMWD would simply use the massive reservoir supply to fight a catastrophic fire. If LVMWD did use the reservoir water directly, it would require a “boil water order” for any water that people drink for a couple of days. There is a health risk in this extreme scenario, and some cost. Directors Polan and Steinhardt indicated that the purpose of a real probability study before approving the tank is simply to aid the board in making an informed, fiduciary decision on which is a more responsible way to spend money and plan water resources.

The easiest way to think about the water reserve design is to think of a glass of water (as described to Westlake Revelations by LVMWD General Manager Pedersen). This is simplified greatly, but it will give you an idea of how this works. The top inch in the glass is for the variation in use during the normal course of a day (some times of day use more water than others). The next couple of inches are to insure that if the water supply gets shut off, there are enough reserves in place until the water comes back on. The bottom couple of inches are in case of a large residential or commercial fire (note: specifically not a brush fire).

According to LVMWD, we’re starting to use enough water that we’ve been dipping into the fire reserves (the bottom layer in the glass). The fire reserves are designed for all of Los Angeles County, not specifically for the western part of the LVMWD service area. And, the fire reserves make no mention of what kind of water — potable (drinkable) or unfiltered. In fact, unfiltered and potable water put out a fire equally as well. Note that the “glass” here does not include the reservoir, it’s only the filtered water stored in tanks. It also does not include any of the water that is in the system “piping” — but that’s normal for water districts to ignore as an additional layer of safety.

To assess the probability of three items like this, you determine the probability of each individual item first. In our case, we wanted to look at not just on a given day, or month, or year — but over a 30 year period. For example, the chances of in 30 years there being a very hot day with peak water usage is effectively 100%. Determining the chances of “delivery failure” or an extraordinarily large commercial fire is a bit trickier.

In the case of delivery failure, there are a variety of reasons delivery could be shut off without notification. (If there’s notification, or the failure happens during certain times of the year, LVMWD’s filtration plant can be turned on and used — just as it already is in the summer time.) To be generous to the tank’s case, we used the chances of a 7.0 earthquake, and ignored the fact that the filtration plant runs throughout the summer period. There’s a good chance that a 7.0 earthquake wouldn’t be enough to shut off the supply, but it’s a good estimation overall. And, earthquake is likely: there’s over an 80% chance a 7.0 earthquake will happen in the next 30 years. And, the filtration plant running continuously during the summer could render a supply shut off irrelevant, eliminating the need for the tank on its own.

Finally, the third component is a commercial fire. In short, the experts we spoke with are having a hard time envisioning the scenario for a large fire (large enough to meet the standard) that would take place on the western part of the LVMWD service area. See below on the explanation about fire.

Director Polan, with Director Steinhardt’s support, was able to convince the LVMWD Board of Directors to investigate how to pursue a probability study. An ad hoc committee was created specifically to look at how to put together a probability study on the actual need for the tank (the simultaneous, triple problem). In September, LVMWD Board President Charlie Caspary unilaterally decided to disband an ad hoc committee. See

Chances of a Large Commercial Fire

As said earlier, LVMWD’s core reason for the proposed 5 million gallon water tank is fire safety. Specifically, that it needs to have a supply of five hours of water at 5000 gallons per minute. Westlake Revelations has been inquiring about the likelihood of a fire of this size. This is difficult information to get as it relies on approximation by experts, but even so, it’s easy to see that the level of probability. Some background first.

A residential fire burns quickly. A house generally burns to the ground in 30-60 minutes. If it were a super large house, it might take 60-90 minutes. During that time, even in a large house fire, it’s unlikely that 5000 gpm would be needed solidly.

The fire standard is *not* for brush fires — they are handled in a different way and water districts are not asked to prepare for them.

A commercial fire can be bigger, and is usually the largest and longest type of fire. To be overly conservative, we looked at the City of Los Angeles — where commercial fires are far more prevalent than in a place like Westlake Village, or Agoura Hills. Note: The City of Los Angeles has a population of 3.85 million. The area in question that is served by LVMWD’s proposed tank (the western portion of LVMWD) has an estimated population of less than 30,000 (less than 1% of LA City’s population size).

In the past 25 years, there appears to have been only a handful of very large-scale fires which would last 5 hours *or* would use 5000 gallons per minute. But, we were unable to find *any* fire in the City of Los Angeles history that lasted 5 hours at 5000 gpm. This standard is such a large fire that even if you took all the fires that happened during the Northridge earthquake (which affected a very large geographic area), it’s unclear if combined there was five hours at 5000 gpm cumulatively. There were fires in the history that exceeded 5000 gpm, but they would only last an hour or two and then the volume of water would be scaled way back. And, there were fires that used more than 5000 gpm, but they were no where near close to long enough to come close to the 1.5 million gallons of 5 hours at 5000 gpm.

The largest of commercial fires that we could find as an example, while not large enough to need this level of water (1.5 million gallons), included a furniture factory facility, a metal working factory, and a refinishing company. Additional examples might include a tire factory, a junkyard, a chemical plant, etc… To our knowledge, there is not a single one of these in the area in question (the western part of LVMWD).

Even a large event — like a natural gas fire triggered by a major earthquake would have a difficult, if not impossible time to make it to this size of fire.

And, even something like a lumberyard, or building supply store, doesn’t have enough fire load to burn for five hours using that amount of water. It would either be extinguished far before five hours, or would burn out (e.g., run out of fuel) after a couple of hours.

It appears to literally be a one in a million chance of a fire this size during any given month. But, to be conservative, we used one in a hundred thousand — an order of magnitude more likely. This gives some chance to a fire of this size in the small local area (again, western portion of LVMWD only), even though we cannot find a single example of a fire this size in all of Los Angeles going back decades.

Calculating the Probability

To calculate the probability, we looked at chances over a 30-year period. Because temperature varies greatly during the year (and water usage along with it), we looked at 5 peak period months of higher water usage (and 7 non-peak). For fire, we assumed the chances don’t change during the course of the year — different seasons bring different risks. And, in the case of supply disruption, we used a 7.0 earthquake as the estimator (earthquake assumptions from

Once we had all the different chances for peak and off peak seasons, we determined the chances of each of the three for a year and then calculated the 30-year chances for year. To get the overall probability for these three unrelated events, we simply multiplied them together for the outcome.

Again, a proper probability analysis can take this rough estimation and calculate in greater detail.