by Nicholas Pinter
A new $37.2 million levee in the town of St. Helena, on the floodplain of the Napa River, has a colorful history and has been stirring local acrimony since its inception. This project illustrates both the attraction of levee protection, in this case protecting a low-income neighborhood (“low income” by Napa standards) hit by disastrous flooding in 1986 and again in 1995, as well as pitfalls of this strategy for long-term risk management. There are clearly positive elements of the St. Helena levee project, but also numerous missteps that have mired the project in dissent and even, opponents argue, threaten to bankrupt the town. With important planning and zoning decisions now pending, the St. Helena levee is a case study for other communities to examine before they consider all of the options for flood-risk management.
After the release of a recent journal article on levee protection, one reporter tried to wake me up with the following opening question – “Prof. Pinter, why do you hate levees?” I laughed in appreciation of a good icebreaker, and then explained how levees, floodwalls, and other structural protections are valuable tools in a diverse toolkit used to manage flood risk. But levees also come with adverse consequences that need to be rigorously evaluated and balanced against their potential benefits. Those negative impacts and the need for careful evaluation are often lost in the rush to build new levees or enlarge existing structures, typically in the aftermath of a damaging flood.
St. Helena is a town of about 6,000 people along the Napa River about 20 miles upstream of the City of Napa. The river there is small, averaging just 46.7 cubic feet per second (cfs) of flow over the duration of record, but subject to flash flooding after intense local rainstorms. Since 1862, 28 major floods have struck the Napa Valley, including 1955, 1963, 1967, 1973, 1978, 1981, 1983, 1986, 1993, 1995, 2002, and 2006. Most of St. Helena is located a safe distance from the river, but starting in the 1970s, the Vineyard Valley area was developed, including 468 homes, apartments, and mobile homes at risk of flooding. Most of these are in what is now mapped by FEMA as the so-called 100-year floodplain. During the five-year review process for development of the mobile home park, local residents report that the issue of flood risk was never raised.
Figure 1 – The Vineyard Valley neighborhood of St. Helena under water during the 1986 flood.
Intense rain events hit the upper Napa Valley in February of 1986 and again in March of 1995, sending the Napa River over its banks and into the new Vineyard Valley development. Damages in each event were estimated at over $50 million in St. Helena alone. During the 1995 flood, over 400 residents of St. Helena were evacuated, and 200 mobile homes in Vineyard Valley were declared a total loss. In both 1986 and 1995, other communities such as Calistoga and Yountville also were hard-hit, including inundation of much of downtown Napa. As described by the Napa Valley Register, “Overnight, the Napa River swelled from a modest tributary of San Francisco Bay to a wrathful Amazon.” The 1986 flood killed three people, forced the evacuation of over 2,500, and caused total damages estimated at $100 million2.
A typical pattern in flood response is that it takes a one-two punch to motivate any large-scale action to address flood risk. The 1995 flood in Napa was that second punch. Surveying the damage, Napa County officials dusted off a flood-control plan initially drafted by the Corps of Engineers but defeated by a voter referendum in the 1970s. The original Corps flood control plan was considerably reworked and, particularly in the City of Napa, many forward-thinking elements were added, including channel habitat preservation and reconnection of the historic floodplain. Flood control was now put to the voters again, in the form of Measure A, a half-percent county sales tax, which was passed in March of 1998 with a two-thirds voter majority. With local cost-share funding secured through the referendum, Napa County began moving forward with the large-scale flood plan in Napa, with smaller portions of the funding apportioned to other communities along the river.
Figure 2 – The positive side of flood control. As part of the St. Helena project, a total of 17 mobile homes were permanently and the flood wall set back from the river.
In St. Helena, plans were drawn up that focused on a large levee and floodwall to protect the Vineyard Valley area, as well as several other elements. These other elements of the project included bank stabilization, removal of a channel constriction, internal stormwater management, and excavation of a storage area outside the levee (a “terrace,” the project calls it) planted in native vegetation. The channel restriction, for example, involved the removal of 17 mobile homes and the associated land closest to the river, which opened that area to better pass flood flows. This bend in the river channel represented a chokepoint during floods, significantly obstructing flow and raising water levels in that area and upstream. Removal of these homesites and augmenting the conveyance capacity of the channel at this chokepoint certainly lowered flood danger by both reducing at-risk infrastructure and by lowering water levels during future floods.
Although the St. Helena project included several forward-thinking elements, it remained much more of an “old school” structural flood-control plan than what was implemented in Napa. The St. Helena plan centered on approximately 2,000 ft of floodwall surrounding the Vineyard Valley mobile home park and another ~2,000 ft of levee that extends from the floodwall to the north and west. As now required of urban levees in California, the system protects to the “200-year” level – that is, designers of the levee and floodwall have affirmed that they will keep out floods up to those with a 0.5% chance of occurring in any year.
On a map the St. Helena levee is relatively small, a tiny fraction of the length of the systems that protect the Central Valley or great alluvial floodplains of the Midwest. But its height is impressive. Standing on the new levee crest above the Napa River, one can almost imagine looking out on the Sacramento or Mississippi River. The levee is also proportionally wide, with a crest 15-20+ feet across, wide enough to meet requirements for much larger rivers and floodplains. Wide enough, in fact, to carry two full lanes of traffic. Indeed, the initial St. Helena planning documents called for a multi-function levee that would also serve as a new roadway. This new road would have crossed the floodplain and the river across a new bridge, connecting St. Helena to the Silverado “wine trail” on the northeast bank. Currently that connection is across the Pope Street bridge, a historic stone structure built in 1894 and so narrow that it is terrifying to cross even when the driver coming the other way hasn’t spent the afternoon sampling the latest vintages. Using flood-control funding to simultaneously serve another goal might not be illegal, but it raises questions about priorities in design details. In this case, the proposed bridge link was rejected after litigation focused on its environmental impacts, but before levee construction. Nevertheless, the width of the St. Helena levee retained the width of a road to nowhere, as well as – importantly – a path that took the levee far wide of the housing it was funded to protect, enclosing a broad swath of floodplain undeveloped and empty except for rows of grapevines. More on this in a moment.
Another interesting aspect of the St. Helena levee is the funding to plan and design the project. The city obtained an $8 million grant from FEMA’s Hazard Mitigation Grant Program (HMGP). HMGP uses a slice of FEMA’s total disaster-related funds to promote “mitigation,” meaning actions taken by communities to lower their overall exposure to future disasters. Flood-mitigation strategies favored by FEMA nationwide include buyouts and relocation of repeatedly flooded homes or elevation of floodplain structures up beyond the reach of future floods.
FEMA carefully monitors HMGP projects to make sure that taxpayer funds are used for their intended purpose. In 2013, when FEMA audited St. Helena, the city was unable to account for a large block of HMGP money. Among project expenditures, the city had spent $1.64 million on lobbying expenses, reportedly including lavish entertainment expenses in Washington aimed at securing Army Corps of Engineers and other federal funding for levee construction. When FEMA reviewed the city’s books, it disallowed $1.9 million in expenses, requiring St. Helena to withdraw this amount from its operational accounts. Further local acrimony ensued because the city repaid the FEMA money without adequately, some argue, disclosing the financial missteps and the resulting impacts on city finances. This repayment resulted in a $1.5 million budget deficit and significant cuts to city services.
The details above are a selection of the complaints voiced by local levee opponents in St. Helena. This dissent contrasts with the broad consensus surrounding the much larger flood-control measures implemented in the City of Napa. Large public projects never reach perfect consensus, but the Napa Project solicited input from a broad range of stakeholders and integrated these opinions in final project designs. In contrast, the St. Helena project continues to roil that community even now that the levee is complete and a done-deal.
The final shoe – probably the most problematic element of the St. Helena flood-control debate – is now dropping, five years after the levee’s completion. The current issue results from the unusual pathway of the levee, which sweeps well north and west of the neighborhoods previously flooded and includes 16.9 acres of undeveloped vineyard land. Undeveloped until now, that is. Just 17 days before the passage of Measure A, the mechanism that funded large-scale flood engineering throughout Napa County, much of the empty floodplain land behind the St. Helena levee was purchased by a private developer. On Sept. 8, 2010 the developer submitted plans to convert the empty agricultural land behind the levee into high-density residential development, including 87 new residential units. This despite an explicit provision in Measure A that “none of these projects are intended or designed to encourage population growth.”
Figure 3 – The Levee Sniff Test
My research group has worked hand-in-hand with floodplain residents to provide them planning guidance and technical expertise. We’ve also studied levee projects that would increase flood heights for neighbors, increase long-term flood risk, or degrade habitats, sometimes in exchange for very limited local flood-protection benefits. After struggling with this balancing act time and again, we’ve proposed a three-step “levee sniff test” to preliminarily judge any proposal. New or enlarged levees and floodwalls may be an appropriate solution if the system is designed to protect floodplain infrastructure that is:
- high in value, and
The most crucial – and often the most politically fraught – of these criteria is the third. The worst case is construction of “field of dreams” levees – new, expensive, often muscular levees surrounding undeveloped or sparsely occupied floodplain land. “Build it and they will come” – field-of-dreams levees are meant to attract new construction and development on the floodplain. Political leaders see expansion of their local tax base, and developers salivate over the prospect of transforming low-cost floodplain land into urban real estate as soon as the new levee “solves the problem of flooding forever.”
By intent or by accident, the St. Helena project created a “field of dreams” levee. The original position of the levee was set by the desire of the city to piggyback a long-sought transportation project onto available flood-control funding. A private developer saw an opportunity for windfall profits and snapped up this floodplain acreage. What remains unclear is why, when plans for the roadway were scrapped, did St. Helena not downsize its levee? Local levee opponents see nefarious motives here, and the optics are not good. In 2008-2009, local residents had proposed several possible levee alignments that hugged the edges of existing neighborhoods that needed protection. These alternatives would have eliminated the “field of dreams” issues, cut construction costs, and would have significantly reduced future flood levels for neighboring properties outside the levee. But these alternatives were rejected at the April 14, 2009 meeting of the City Council. Why? Perhaps simply a result of frustration at the grassroots opposition and the desire to move forward with the existing blueprint, however much conditions had changed since those plans were drawn up.
“Residual Risk – the small but real possibility that a levee may fail or be overtopped during future floods.”
So past mistakes aside, why not fill in every square inch of land behind a levee once you’ve got one? The problem with levees, and with field-of-dreams levees above all, is that no type of structural flood protection can ever reduce the flood risk to zero. Even the strongest levee or floodwall carries a “residual risk” of failure or overtopping during large floods. Furthermore, having flood waters pour over or burst through a tall barrier turns the gradual process of floodplain inundation into a much more violent, dangerous, and damaging event. The history of occupation of U.S. floodplains is a history of levee failures, with headlines almost every year bringing footage of disastrous levee failures somewhere in the country. One levee system on the Mississippi River, not an atypical one, was breached 14 times through its history, only to be patched and raised each time, in the process driving up flood levels for the next event and contributing to the next failure. Expert after expert, blue ribbon panel after panel, has implored engineers, local leaders, and floodplain residents to not forget about residual risk. Brig. Gen. (Ret.) Gerald Galloway, former commander of the US Army Corps of Engineers puts it this way, “Let no one believe that because you are behind a levee, you are safe” (Galloway, 2005). The point has also been made more bluntly, “There are two kinds of levees … [t]hose that have failed and those that will fail” (Martindale and Osman, 2010).
“Developers and home builders may be gone, but residents of Natomas will be there and are the ones who are rolling the residual-risk dice year after year.”
Neglecting residual risk of flooding behind levees creates a false sense of security for residents and a lunatic calculus in which the local community can find itself with far greater long-term flood risk than before the levee was built.
California has made these mistakes before. In fact, the leading example of a “field of dreams” levee, discussed in floodplain management circles around the world, is Natomas on the northwest side of Sacramento. The Natomas Basin is ringed by 42 miles of levees, encompassing approximately 53,000 acres of the floodplains of the Sacramento and American Rivers. This land was sparsely populated until roughly 1990. Despite a tangled history, the Natomas system was built and they did come. On-going levee construction, particularly improvements initiated in 1992, ushered in a wave of development. As of 2010, Natomas was home to over 22,800 structures, the large majority of them residential units and home to more than 90,000 people. The City of Sacramento currently plans to permit up to 1,000 additional single-family homes per year, 500 multifamily residential units per year, and unlimited industrial, commercial, and retail construction behind this levee. Today, total property in Natomas at risk of flood damage, should the levee fail or be overtopped, exceeds $8.5 billion, with flood depths that could reach 25 feet. Heroic efforts are now underway by the Sacramento Area Flood Control Agency, the California Central Valley Flood Protection Board, and the Corps of Engineers to patch deficiencies in the Natomas levee and raise the overall protection provided by the system. When that huge – and hugely expensive – effort is complete, the State of California, the Corps, and FEMA will sign off that they have reduced the statistical chance of catastrophic flooding in Natomas to 0.5% per year (less than once every 200 years, on average). That is a high level of protection by US standards, and engineers and leaders who are implementing the project will be justifiably proud of their accomplishment.
But … our country’s 400+ year history with levees also tells us that such claims of almost complete flood protection have often been repeated, and almost always were over-stated. Lots of reasons why this has happened in the past, and many of those reasons hold true for the future. But even at face value, a 0.5% exceedance probability for a system now protecting $8.5 billion in property translates – crudely – into up to $42.5 million in risk each year. If those 22,000+ homes and other structures had already been there before the levee (concentrated, high-value, and preexisting), then this annual $42.5 million in residual risk would have been a necessary tradeoff. But because Natomas was a field-of-dreams levee built intentionally to create new floodplain infrastructure, the large majority of this $42.5 million/year is all-new and added risk.
Who bears this residual risk if (some would say “when”) California’s drought turns to flood and hands us a “>200-year flood?” Or a lesser flood that finds a chink somewhere in the 42-mile long, gopher-prone armor that surrounds Natomas? Developers and home builders may be gone, but residents of Natomas will be there and are the ones who are rolling the residual-risk dice year after year. So too are all US taxpayers, since we bankroll disaster relief. Federal expenditures for disaster relief have averaged ~$35 billion per year in recent years, equivalent to about $400 per year per US household.
Residual risk also can be borne by the governmental agencies that facilitate “field of dreams” levees. Legal liability came to roost in California in 2003, when the Court of Appeals found the state liable for $500 million in damages resulting from a levee break on the Yuba River in February of 1986, the same storms that flooded St. Helena. The case was Paterno v. State of California, and although the state neither built nor maintained the levee that gave way, the court found that it was responsible for inadequate maintenance and management of the system. And liability extends to counties and towns as well; for example, California Assembly Bill 80 made local jurisdictions liable for “property damage caused by a flood [if they] … unreasonably approv[e] new development in a previously undeveloped area.” In the Paterno case, the financial hit of that $500 million decision shook the financial structure of the entire state, such that California was forced to borrow funds from Merrill Lynch in order to pay the claims against it.
So where does St. Helena and its levee fit into this broader picture? This levee and the new construction proposed behind it are tiny compared to what has gone onto Sacramento area floodplains, St. Louis floodplains, Houston floodplains, and others. But the principles are the same, and the potential consequences for a small community like St. Helena are proportionally even more threatening than for a large city. The St. Helena Vineyard Valley levee was built to protect residents who had the misfortune of living in a neighborhood built on the edge of the Napa River and far into the floodplain, a neighborhood that should not have built in the first place. That was certainly mistake number one. But after repeated and devastating floods, those residents were justified in seeking a mitigation solution. The second crucial misstep occurred in 2009, when the city approved a levee map that enclosed a large area of floodplain empty except for rows of grapevines. Whether by nefarious purpose or just fatigue, St. Helena opened the door to the crucial and difficult decision now in front of them – whether to bow to the pressure and temptation of approving development of that empty land behind their levee.
Figure 4 – View from Vineyard Valley Levee access road, towards the east, showing current conditions of the proposed development area. Figure IV.A-3a in Hunter Valley Subdivision Project, Draft EIR, Dates 5/29/2012.
One concern certainly on the minds of St. Helena city leaders is the threat of litigation if they deny the development plans in front of them. When such cases go to court, plaintiff lawyers typically cite the Fifth Amendment of the US Constitution, which states that “private property [shall not] be taken for public use, without just compensation.” The legal claim is that refusal to allow development of the owner’s land constitutes a “takings” of the value of that land. The St. Helena situation is typical of many examples where the legal threat of a takings claim is used to thwart zoning or planning to manage flood risk and other natural hazards. Developers and their lawyers argue that zoning to prevent development is unjustified, arbitrary, and an infringement of the rights of landowners. On the contrary. The St. Helena developer purchased agricultural land in a floodplain at a price consistent with those conditions. Despite the construction of a levee at taxpayer expense surrounding that land, it remains a floodplain. The limitation on building expensive homes and apartments on that floodplain is imposed by nature, not the City of St. Helena.
Fear of lawsuits and potential liability is a tangible pressure on a town like St. Helena. But in considering development of floodplain land behind a field-of-dreams levee, these communities are perhaps looking at the lesser threat. Bowing to the will of the developer dodges a likely “takings” lawsuit, but loads St. Helena with new residual risk – risk that is small as an annual percentage, but large in total magnitude. The mean value of a detached home is St. Helena is $640,000 and an attached unit is $401,000. Given current plans for 51 new single-family homes, 11 granny units, 24 multi-family units5, and a typical value of 50% content damages for flood-loss estimates, these 87 new units could suffer $49 million in damages if that area were inundated. After the 1995 flood, residents of Vineyard Valley sued the City, and residents of the proposed new development could take the same action. The 2003 Paterno decision against California amounted to just 0.5% of the state’s total budget that year. For comparison, the City of St. Helena, with revenue to its general fund of $10.5 million in FY 2015/16, could face a financial hit totaling 469% of its annual resources. In other words, the legal liability from the proposed new development could bankrupt the city nearly five times over.
“The legal liability from the proposed new development could bankrupt the city nearly five times over”
St. Helena was founded in the mid-1800s, near enough the Napa River to gain the benefits of that proximity, but up and off the floodplain. However beginning in the 1970s, the town suffered the same magnetic pull as countless other communities across the US – succumbing to pressures for growth and the attraction of empty land ever closer to the river. Located deep in the Napa River floodplain, the Vineyard Valley neighborhood was reviewed and approved, reportedly without even discussing the possibility of flooding. This was an oversight for which residents paid dearly in 1986 and again in 1995. Mistake followed mistake: land that never should have been developed, an expensive levee that has already sent the town into financial distress and political discord, and a levee map that somehow enclosed much more floodplain land than the neighborhood it was designed to protect. St. Helena now has a last chance to bring this cascade to an end and avoid construction of 87 new homes behind their field-of-dreams levee. Floodplain managers, river scientists, and prudent leaders worldwide counsel that levee protection is never absolute, and that even the tallest and strongest wall comes with an inherent residual risk, a small but real possibility that it may fail or be overwhelmed during future floods. For concentrated and existing floodplain infrastructure, levees can be the right solution, and residual risk preferable to no action at all. But towns like St. Helena should understand that levee protection brings a range of consequences, including the possibility of multiplying, not diminishing, the long-term risk of damage and potentially even loss of life in their community.
Nicholas Pinter is a professor in the Department of Earth and Planetary Sciences and an affiliate of the Center for Watershed Sciences. This article was developed with assistance from Megan Nguyen, a junior researcher at the Center for Watershed Sciences.
 Report to the St. Helena City Council, Council Meeting of January 26, 2016
 City of Napa. 2009. Hazard Mitigation Plan, 2009 update.
 City of St. Helena, 2007. Project Summary: City of St. Helena Flood Protection and Flood Corridor Restoration Project
 e.g., “Enhanced Minimum Plan” dated 7/2003; City of St. Helena Staff Report dated 4/14/2009
 Hunter Subdivision Project, Draft EIR, dated 5/29/2012 (Sch: 2012 032048), Available from http://cityofsthelena.org/sites/ default/files/Hunter%20Public%20Review%20Draft%20EIR_%20052912.pdf
 2010 Census data; North Natomas + South Natomas.
 City of Sacramento, 2015. Fact Sheet: Next steps for Natomas with a revised flood map, dated 3/24/15.
 US Army Corps of Engineers, A. Kirchner, 2010, Road to a Chief’s Report: American River Watershed [Common Features] Project, Natomas Basin, Sacramento and Sutter Counties, California.
 Disastrous Spending: Federal Disaster-Relief Expenditures Rise amid More Extreme Weather. Center for American Progress, 2016.
 California Senate Judiciary Committee, 2007. AB 70: Bill Analysis. Available from http://www.leginfo.ca.gov/pub/07-08/bill/asm/ab_0051-0100/ab_70_cfa_20070711_145337 _sen_comm.html
 Vogel, N., 2005. A wave of relief after 1986 flood. Los Angeles Times, Aug. 15, 2005. (Available from articles.latimes.com/2005/aug/15/local/me-flood15)
 2013 data from http://www.city-data.com.
Pinter, Nicholas et al. 2016 Modeling residual flood risk behind levees, Upper Mississipi River, USA. Environmental Science & Policy. 58 131-140