City of Modesto v. Dow Chem. Co.

Rivera, J.

*135I. INTRODUCTION

In late 1998, the City of Modesto (the City), the City of Modesto Sewer District No. 1 (the Sewer District) and the Modesto Redevelopment Agency (the RDA) sued various retail dry cleaning businesses (dry cleaners) operating in Modesto together with *767the manufacturers of dry cleaning equipment used at those dry cleaners, and the manufacturers and distributors of dry cleaning solvent. Plaintiffs alleged that defendants had caused the City's groundwater, sewer system and easements, and the soil of property located within the project area of the RDA, to become contaminated with perchloroethylene (PCE), a "toxic chlorinated solvent[ ]." Plaintiffs sought recovery for the past, present and future costs of investigation and remediation of the contamination at numerous sites under multiple legal theories.

This action has engendered nearly 14 years of litigation, including three detours to this court, and five trial phases. A final judgment was entered in November of 2011, and an amended judgment in May 2012. To the extent it can be summarized in one sentence, the judgment awarded plaintiffs damages with respect to three dry cleaning sites, including an award of punitive *136damages against three defendants; as to all other claims, judgment was entered in favor of defendants.

An issue that has been central to this litigation from the outset is the interpretation and application of the Polanco Redevelopment Act ( Health & Saf. Code, § 33459 et seq. ) (the Polanco Act),¹ which, in essence, authorized redevelopment agencies to remediate contamination found in property, including private property, located in a redevelopment project area, and to recover from the "responsible parties" the costs of the cleanup. Early in the case, this division issued a decision, on a petition for writ of mandate, providing our construction of that law. Specifically, we reversed the trial court's summary adjudication of the Polanco Act claims, concluding that the trial court erred in finding, as a matter of law, that defendants could not be "responsible parties" under the Act based on the facts put forward by the plaintiffs. ( City of Modesto Redevelopment Agency v. Superior Court (2004) 119 Cal.App.4th 28, 13 Cal.Rptr.3d 865 [ Modesto I ].)

Thereafter, the Polanco Act claims were tried two times, with respect to different contaminated sites, before two different judges, with different results. In the second proceeding, the trial court concluded that Modesto I implied a special causation standard was applicable to the Polanco Act claims. In the published portion of this opinion we hold that no special causation standard applies and we will, accordingly, vacate the trial court's ruling and order on the Phase IV Polanco Act claims.

In the unpublished portion of the opinion we address the remaining issues on appeal. We will vacate the trial court's pretrial ruling with respect to plaintiffs' nuisance claims; the punitive damages award against defendant, the Dow Chemical Company (Dow) in Phase I; and the trial court's directed verdict on grounds of no present injury as to various sites in Phase III. Our determinations with respect to the statute of limitations, the denial of equitable relief, the amended judgment, the prevailing parties and the allocation of settlement credits will be described in the course of the opinion.

II. GENERAL BACKGROUND

We begin by providing some general technical information about the characteristics of PCE, how it can contaminate groundwater, how it is used and reused in dry cleaning equipment, and how it was released into the environment. This is not a comprehensive discussion of the evidence but is intended only to supply context for our discussion of the issues on this appeal.

*768*137A. Characteristics of Perchloroethylene

Perchloroethylene, also known as tetrachloroethylene, is a molecule containing chlorine atoms and carbon atoms. It is also characterized as a "volatile halogenated organic compound," a "halogenated hydrocarbon", a "chlorinated solvent" or a "chlorinated hydrocarbon". As shorthand, it is referred to as "perc" or PCE. All chlorinated hydrocarbons, like all solvents other than water, are "toxic." In 1978, the National Institute for Occupational Safety Hazards (NIOSH) recommended that PCE be handled as if it were a human carcinogen. In 1980 the State of California began regulating PCE as a hazardous waste. In 1984, when the Resource Conservation Recovery Administration (RCRA) was reauthorized, its regulations brought "small dry cleaners" under the same requirements as major hazardous waste sources, with respect to PCE.

The California Department of Health Services has set the maximum contamination level (MCL) for PCE in drinking water at 5 parts per billion based on its finding that PCE potentially causes cancer in humans. Applying this standard, if one cup of PCE were completely dissolved in water, it could contaminate 24 million gallons of groundwater. There are also regulations to prevent migration of PCE vapors in concentrations that could cause cancer. All parties agree that the applicable regulatory standards are not arbitrary and address genuine public health risks.

PCE is a colorless liquid, and is therefore difficult to see once released into soil. It is a cleaning solvent used by dry cleaners and also in degreasing operations. Because in its pure form it is a dense non-aqueous phase liquid (DNAPL), it is heavier than water and so when placed in water it will sink and sit below the water. This is distinguished from a light non-aqueous phase liquid (LNAPL), such as gasoline, that is lighter than water and will therefore float on top. PCE also has lower viscosity (internal friction) than water and so it is very mobile and can move quickly to penetrate, for example, small cracks or joints in concrete. PCE does not readily dissolve in water-thus, "non-aqueous"-although it will dissolve very slowly over time. PCE is also quite volatile, meaning it will quickly become a gas when it is heated or released into soil where it mixes with the soil gas.

As is explained below, PCE is particularly "persistent" and "long lived" compared to other contaminants, making it extremely difficult to accomplish complete remediation.

B. The Flow and Transport of PCE in a Groundwater System

The City of Modesto uses groundwater as a primary source of its drinking water supply. The hydrologic cycle for groundwater is fairly straightforward. The *138rain falls and sinks into the "soil zone" which is the layer near the surface where plants and trees take it up with their roots. The rest of the water continues to move downward through the soil into the "vadose zone," in which there are various types of soils-fine grain such as clay or silt and coarse grain such as sand or gravel. Some of the water may "perch" on the fine grain materials but other water will flow through the coarser soil media and recharge the water table, which is at the top of the "saturated zone" from which ground water is pumped by wells. Between the vadose and saturated zones is a "capillary fringe." This space can hold "quite a bit of liquid," and the porous media can pull water up from the water table. *769As we have noted, when PCE is released into the ground it can volatilize and mix with the soil gases. Where there is a very high concentration of PCE in the soil the PCE gas, which is denser than air, will tend to sink. If it reaches all the way to the capillary fringe it can then recondense on the water table and begin to dissolve into the groundwater.

When released into the ground, PCE can also remain in liquid form and will sink through the subsurface to the water table. Unlike gasoline-which is a LNAPL and will tend to float on top of the water table-PCE is a DNAPL and so will continue to move past the water table and through the saturated zone. Because it dissolves into water only very slowly, it creates a long-term continuing source of contamination to any groundwater that comes in contact with it.

Small releases of pure PCE will tend to penetrate only into the vadose zone because it gets "pulled apart" as both a gas and a liquid and will eventually get trapped in the soil pores-much like a small amount of coffee will be absorbed into and trapped in a sugar cube. When more PCE is added, however-either a larger release or repeated releases-the PCE breaks out and penetrates deeper. Even if it pools or perches on the fine-grained soil layers, it can still stair-step down through the subsurface and reach the capillary fringe and-if there is enough PCE-the water table. If PCE is being cleaned out of the groundwater, the PCE that pools or has been trapped above the water table will continue to move into the water table-a process called "back diffusion"-and thus is another long term source of groundwater contamination. PCE that has been trapped in the subsurface can also move downward during a "recharge event;" the PCE can dissolve into the rainwater and be released from the vadose zone, and is yet another long-term source of contamination.

PCE that pools on top of fine-grained layers will stop moving vertically but then will begin to move in other directions due to differing soil strata and "fine textural changes" in the sand; as a result, PCE will have a very *139complicated pattern of distribution, making it more difficult to locate. Also, because DNAPL's can be trapped in various ways, and then dissolve and move over time, they are extremely difficult to clean up. There are sites in the United States where they have been trying to clean up PCE for decades.

C. The Use, Reuse and Release of PCE and PCE Residue at Dry Cleaners

As we have described, PCE is particularly "persistent" and "long lived" compared to other compounds, making it difficult to remediate but advantageous for industrial use because it can be reused after being distilled back to its pure state. Dry cleaning equipment therefore has the capacity to do that. Here we provide an elementary description of how the various machines work.

From the 1950's to the 1970's, dry cleaners used what is now referred to as "first generation" equipment. In that iteration, the washer and the dryer were separate machines. In this equipment PCE was continuously pumped into, and drained out of, the washer from a tank on the bottom of the machine into the basket containing the clothing. As the solvent drained out it was sent through a filter, to clean out the solids, dyes and soil being removed from garments, before it was pumped back into *770the washing basket. After the cleaning cycle was complete, the washer entered a spin cycle to remove and drain as much solvent as possible back into the tank. The clothing was then moved to the "reclaiming dryer." This dryer did not vent to the air during the drying cycle; the vapors were kept in the machine in order to recover the solvent. The dryer also had a lint trap. As will be explained, the vapor was then cooled using water coils to recondense the solvent back to liquid form.

The "second generation" machines were used from the 1970's to the mid-1980's. This equipment operated in the same way, but combined the washing and drying functions into one "dry-to-dry" machine. Also, after reclaiming most of the vapor, some residual vapors were vented into the air at end of the dry cycle. An operator, however, could add a "sniffer" (carbon adsorber) to capture the vapors. These were very large and contained perhaps 100 pounds of charcoal or carbon in which the vapors were caught.

The "third generation" machines were used from the mid-1980's to the mid-1990's. These were also dry-to-dry machines, but used refrigeration coils instead of water to cool the vapor, which was far more efficient in turning vapors to liquid for recovery. The fourth generation machines added a built-in carbon adsorber to filter out the last traces of PCE vapors before the door to the machine was opened.

Reclaiming dryers (or "reclaimers") operate with hot air, so the solvent on the clothing vaporizes. Those vapors are continually blown out of the *140reclaimer by a fan, through the lint filter. The vapors enter a pipe with cooling coils, where the solvent turns to liquid and drips off the coils into a pipe leading to the water separator.

The water separator is a device by which any water-even just from humidity in the air-is removed from the recaptured solvent. All used solvent, no matter from where it is reclaimed or distilled, must go through the water separator before it is reused. The water separator operates by gravity. Because PCE is heavier than water, it will settle to the bottom of the separator. It is then siphoned out by a drain that is piped back to the tank. The water, called "contact water" or "separator wastewater" is drained out through a different pipe located higher in the separator, and was (during the pertinent time period) sent either down a drain or into a pail, depending on how the machine had been set up. All contact water has a certain amount of PCE left in it, because the separation is "imperfect."

As noted, some dry cleaners added a sniffer, or carbon adsorber, to collect solvent vapors. To release the solvent from the sniffer, live steam would be applied to the charcoal bed to release the solvent as vapor; the steam and vapors were then vented to the pipe with the cooling coils where they would condense and go to the water separator. This process creates significantly more separator wastewater than other reclaiming processes because of the large amount of steam.

The final major piece of equipment is the still, which is used to clean the solvent. All solvent is periodically sent to the still, because it will accumulate residual oil, grease, wax, detergent and other impurities. The solvent is heated with steam coils until it begins to vaporize. The vapors are sent through pipes with cooling coils and then to the water separator. The residue, that is, what remains behind after the PCE is vaporized, is called "still bottom residue" or "muck."

*771A still is also used to extract PCE from the filters in the machines that capture the solids, dyes and soils. Historically, dry cleaners used diatomaceous earth (a powder) to filter the PCE, but later they began to use cartridges filled with carbon or carbon and activated clay; some had pleated cellulose on the outside. The still had mechanical paddles to continuously stir the mixture to release the PCE. The residue, or "muck" was then shoveled into a container, and disposed of as waste. This residue contained about 20 percent to 25 percent PCE. A similar process was used for the cartridges. When the cartridge needed to be changed it was removed and drained for 24 hours; that liquid went into the still and was processed leaving a residue like very heavy heating oil, which was drained out through a valve into a pail. In the first *141generation equipment the PCE content left in the residue could be as high as 50 percent. In the new equipment (being sold in the early 2000's) there is as little as 10 percent.

Even after draining the filter for 24 hours there could be as much as a gallon and a half of PCE remaining in the filter. At some point prior to 1984, a process was developed by which, like cleaning out a sniffer, high pressure steam was used to release the remaining PCE from the filter cartridge, although it could never remove 100 percent. This was accomplished by a "solvent recovery system" or SRS. The cartridges were placed into a tank and steam was released into the tank to vaporize the PCE which, again, goes to the water separator. This involves a great deal of steam, and creates a "tremendous" amount of contact water-up to 20-25 gallons.

In very general terms, PCE releases into the environment occurred as a result of equipment failures or leaks during operation or maintenance of the machine or the water separator, leaks in joints or valves due to vibration of equipment, and leaks or spills during solvent delivery or transfers which are flushed to the outdoors, or which permeate the concrete floors, or seep through the cracks or joints in the floors. PCE releases also occurred as a result of tossing filter cartridges into the garbage, tossing muck into the trash or into dumpsters, or "out the back door"; and by PCE disposal practices which have been described as "dumping" or "back lot burial." PCE also entered the soil by releases of separator water on the ground or down the drain and into the sewer system, thence out of sewer pipes into the ground.

In addition, those servicing the dry cleaners on behalf of the manufacturers or distributors performed tests to check on the amount of detergent in the solvent. The dry cleaners did not want that liquid to be put back in the machines because it used a strong dye, so as a practical matter the test liquid, which contained about 25 percent PCE, was poured down the drain.

D. Contamination of Soils, Groundwater and Wells in Modesto

Defendants do not dispute that the soils at many dry cleaning sites in Modesto are contaminated by PCE. It is also not the subject of serious dispute that this contamination has dispersed through the subsurface and into the groundwater at many sites. PCE contamination can enter a municipal supply well by migrating with the groundwater generally according to the natural gradient, but when it is near a well, it will come within the "cone of draw" and will be captured and pumped into the well.

The residents of Modesto rely on groundwater wells to produce most of *772their potable water. The City is served by a network of 90 wells designed to *142serve specific neighborhoods, and operating interdependently. If a well is taken off-line, nearby wells are required to pump at greater capacity and the contamination plume is likely to migrate to those wells.

PCE has been detected in four of the City's wells, and two were removed from service after exceeding the state-mandated MCL. The parties dispute whether and when the third and fourth wells (or additional wells) will exceed the MCL for PCE.

This action was filed as a result of the PCE contamination from the dry cleaning sites in Modesto.

III. PROCEDURAL HISTORY

A. Overview^**

B. The Parties to the Appeal

The City of Modesto is a plaintiff and the principal appellant. The RDA also filed a separate action, which was consolidated with the City's action. In 2012, however, the legislature effectively dissolved all redevelopment agencies. (§ 34172.) The City became the RDA's "successor agency" by operation of law (§ 34173), and pursues the RDA's appeal, together with the City Attorney of Modesto who, representing the People of the State of California, seeks a nuisance injunction. The Modesto Sewer District No. 1 was also a plaintiff and a cross-defendant but is involved in this appeal only with respect to the cost award entered against it. For ease of reference we will refer to all plaintiffs as "the City."

The operative complaints named 28 defendants. Along the way, however, most of the defendants settled, and there now remain only five active participants. They are: two dry cleaning establishments, Modesto Steam Laundry (MSL) and Estate of Shantilal Jamnadas, dba Halford's Cleaners (Halford's); two PCE manufacturers, the Dow Chemical Company (Dow) and Axiall Corporation, the successor in interest to PPG Industries, Inc. (PPG); and one PCE distributor, R.R. Street & Company (Street), which also manufactured some equipment used by the dry cleaners. Dow is the sole cross-appellant on the issue of punitive damages.

C.-L.^***

*143IV. DISCUSSION

A. Modesto I and the Causation Standard

Plaintiffs contend that in the Phase IV Polanco Act bench trial, the court erred in applying an exceptionally stringent standard of causation based on its erroneous interpretation of Modesto I . We agree and our analysis follows.

1. The Precise Issue on Appeal^†

2. The Polanco Act

We begin our discussion with a summary of the Polanco Act and then provide a more detailed description of how that law was interpreted in Modesto I.

The trial court aptly summarized the Act's core provisions. " 'The Polanco Act involves cleanup of the release of hazardous substances in the context of a redevelopment project.' [Citation.] Subject to certain statutory conditions, the Act authorizes *773a redevelopment agency to 'take any actions that the agency determines are necessary ... to remedy or remove a release of hazardous substances on, under, or from property within a project area, whether the agency owns the property or not....' [Citation.] If an agency takes such action, 'any responsible party or parties shall be liable to the [ ] agency for the costs incurred in the action.' [Citation.] An action for cost recovery under the Act 'is in addition to, and is not to be construed as restricting, any other cause of action available to a redevelopment agency.' [Citation.]" Thus, under the Polanco Act, a redevelopment agency is entitled to recover from "any responsible party or parties" the costs it incurs "to remedy or remove, or to require others to remedy or remove ... a release of hazardous substance" including "compelling a responsible party through a civil action, to remedy or remove a release of hazardous substance." (§ 33459.4, subd. (a).)

The term "responsible party" is defined by reference to two other laws. First, the Act incorporates a provision of the HSAA (§ 25323.5, subd. (a)(1)), which, in turn, incorporates CERCLA's definition of "covered persons" for purposes of liability ( *14442 U.S.C. § 9607(a) ).⁶ The City does not contend defendants (other than the dry cleaners themselves) fall within that definition. Second, the statute incorporates section 13304, subdivision (a) of the Water Code which provides, in pertinent part, that any person who "has caused or permitted, causes or permits, or threatens to cause or permit any waste to be discharged or deposited where it is, or probably will be, discharged into the waters of the state and creates, or threatens to create, a condition of pollution or nuisance" shall be responsible for remedial action. It is this definition that was the subject of our prior opinion in this matter.

3. The Modesto IDecision

Early on in this case, the manufacturer defendants filed a motion for summary adjudication of the Polanco Act and negligence per se claims. They argued they were not "responsible parties" under the Polanco Act because they neither themselves discharged the PCE nor did they control any site where the discharges occurred. The trial court granted the motion, ruling that the manufacturers were not "responsible parties" under the statute because in order for a party to "cause" a discharge "you have to have some sort of physical control or the ability to stop it from happening." ( Modesto I , supra , 119 Cal.App.4th at p. 36, 13 Cal.Rptr.3d 865.) Plaintiffs filed a petition for writ of mandate in this court, which we granted because we disagreed with the trial court's conclusion that only those who physically engage in a discharge or who control the waste disposal activities of others are liable under the Polanco Act.

In the decision, we first rehearsed the statutory language. As we have already *774described, the Polanco Act defines "responsible parties" by reference to Water Code section 13304, which provides: "A person ... who has caused or permitted, causes or permits, or threatens to cause or permit any waste to be discharged or deposited where it is, or probably will be, discharged into the waters of the state and creates, or threatens to create, a condition of pollution or nuisance, shall ... clean up the waste or ... take other necessary remedial action ...." We therefore began our analysis by asking whether the word "cause" refers to "a party who was directly involved with a discharge, to anyone whose actions were a substantial factor in causing *145the discharge, or even, as city argued below, to anyone who places a hazardous substance into the chain of commerce." ( Modesto I, supra, 119 Cal.App.4th at p. 37, 13 Cal.Rptr.3d 865.)

To answer that question we noted, first, that environmental legislation by which government exercises its traditional power to regulate public nuisances should be construed in light of common law principles bearing on nuisance, citing Leslie Salt Co. v. San Francisco Bay Conservation etc. Com. (1984) 153 Cal.App.3d 605, 200 Cal.Rptr. 575. Determining that the Porter-Cologne Water Quality Control Act ( Wat. Code § 13000 et seq. ) was in fact such a legislative scheme ( Modesto I , supra , 119 Cal.App.4th at pp. 37-38, 13 Cal.Rptr.3d 865 ), we identified certain principles which would govern our construction of the statute.

The first principle, which has "long been the law in California," is that " ' "[n]ot only is the party who maintains the nuisance liable but also the party or parties who create or assist in its creation are responsible for the ensuing damages." ' [Citation.]" ( Modesto I , supra , 119 Cal.App.4th at p. 38, 13 Cal.Rptr.3d 865.) Thus, "liability for nuisance does not hinge on whether the defendant owns, possesses or controls the property, nor on whether he is in a position to abate the nuisance; the critical question is whether the defendant created or assisted in the creation of the nuisance. [Citation.]" ( Ibid. )

Our second principle was a limiting one. We concluded that, "[w]hile liability for nuisance is broad, ... it is not unlimited" and the " City of San Diego [v. U.S. Gypsum Co. (1994) 30 Cal.App.4th 575 [35 Cal.Rptr.2d 876] ] established one important limitation." ( Modesto I , supra , 119 Cal.App.4th at p. 39, 13 Cal.Rptr.3d 865, fn. omitted.) In that case the city sued the manufacturers, distributors and suppliers of asbestos-containing building materials that had contaminated the city's buildings, seeking recovery, inter alia, under a nuisance theory for the costs of abatement. The court concluded that the city could not maintain an action based on nuisance where it is seeking recovery for a defective product, because it would convert almost every products liability action into a nuisance claim. ( Ibid. ) The court affirmed the summary adjudication in favor of the defendants because it was " 'a products liability action in the guise of a nuisance action' [citation]." ( Ibid. ) We agreed with that conclusion, expressing the view that the law of nuisance "is not intended to serve as a surrogate for ordinary products liability." ( Ibid . fn. omitted.)

We then proceeded to the next question: Whether, in this case, "the Polanco Act claims fall within the realm of nuisance or of products liability;" that is, "has city presented evidence that the defendants assisted in the creation of a nuisance, or only that they produced or supplied defective products?" ( *775Modesto I , supra , 119 Cal.App.4th at pp. 39-40, 13 Cal.Rptr.3d 865 [emphasis added].) We looked *146first to Selma Pressure Treating Co. v. Osmose Wood Preserving Co. (1990) 221 Cal.App.3d 1601, 271 Cal.Rptr. 596 ( Selma ), disapproved on other grounds in Johnson v. American Standard, Inc. (2008) 43 Cal.4th 56, 70, 74 Cal.Rptr.3d 108, 179 P.3d 905. There, the State and the Regional Water Quality Control Board sued the operators of a wood treatment facility, alleging they improperly disposed of hazardous waste. The plaintiff sought, among other things, damages flowing from a nuisance. ( Selma, supra, 221 Cal.App.3d at p. 1606, 271 Cal.Rptr. 596.) The defendants cross-complained against the company that designed the wood treatment technique, installed the equipment, provided training and made recommendations on operations that resulted in the wood-treating chemicals being deposited into soil overlying an aquifer. Other cross-defendants were chemical suppliers that provided "assistance and advice" and knew or should have known that the disposal could threaten the safety of the water supply, but failed to warn of those risks.⁷ ( Id. at pp. 1607, 1609, 271 Cal.Rptr. 596.) The court concluded that both were potentially liable as persons who created or assisted in creating the nuisance. The designer and installer could be liable because it had direct involvement in creating the disposal system by which the waste products were discharged into an unlined dirt pond, which could threaten the water supply. ( Id. at p. 1620, 271 Cal.Rptr. 596.) The court also held that the chemical companies could be held liable if plaintiffs proved their allegations that "the [operators] were foreseeable users; disposal of the chemical residue was a foreseeable use of the product; the chemical companies knew or should have known of the dangers of improper disposal of the chemicals; the owners did not know of those dangers; the companies failed to warn of the dangers; and that failure to warn was a substantial factor in causing the damage. ( Selma, supra, 221 Cal.App.3d at pp. 1621-1624 [271 Cal.Rptr. 596].)" ( Modesto I, supra, 119 Cal.App.4th at p. 40, 13 Cal.Rptr.3d 865.)

We agreed with the first conclusion-"that those who create or assist in creating a system that causes hazardous wastes to be disposed of improperly, or who instruct users to dispose of wastes improperly, can be liable under the law of nuisance." ( Modesto I , supra , 119 Cal.App.4th at pp. 40-41, 13 Cal.Rptr.3d 865.) Citing plaintiffs' evidence that some of the defendants instructed the dry cleaners to set up their equipment in a way that water containing PCE would be discharged directly into drains and sewers, and other defendants "gave dry cleaners instructions to dispose of spilled PERC on or in the ground," we concluded, "these kinds of affirmative acts or instructions could support a finding that those defendants assisted in creating a nuisance, and therefore would defeat a summary adjudication motion on the Polanco Act cause of action." ( Id. at p. 41, 13 Cal.Rptr.3d 865.)

*147Defendants had argued that liability should be limited to those who "controlled either the discharge activity or the premises where the discharge occurred" citing to a number of State Water Resources Control Board decisions. We rejected that argument because the Board's decisions did not address "the responsibility of a party that instructs users to dispose of hazardous *776wastes in an unsafe manner or a party that creates a system that would result in improper disposal of hazardous wastes." ( Modesto I , supra , 119 Cal.App.4th at p. 41, 13 Cal.Rptr.3d 865.) We also pointed out that the State Board has concluded that even a relatively minor contribution to a discharge may support a finding of responsibility, citing In reCounty of San Diego (Order No. WQ 96-2, Feb. 22, 1996) 1996 Cal. ENV LEXIS at p. *14 (Cal.St.Wat.Res.Bd.).

At the other end of the spectrum, the City had argued that liability for nuisance can be fixed by proving only that defendants manufactured and sold solvents to dry cleaners with knowledge of the hazards of those substances and without alerting the dry cleaners to proper methods of disposal. ( Modesto I , supra , 119 Cal.App.4th at p. 42, 13 Cal.Rptr.3d 865.) We rejected that contention as well. We reasoned that "failure to warn [is] not an activity directly connected with the disposal of solvents. In our view, such behavior is analogous to the manufacture, distribution, and supplying of asbestos-containing materials in City of San Diego [v. U.S. Gypsum Co. (1994) 30 Cal.App.4th 575 [35 Cal.Rptr.2d 876] ]; it does not fall within the context of nuisance, but is better analyzed through the law of negligence or products liability, which have well-developed precedents to determine liability for failure to warn. [Citations.]" ( Ibid. )

We, accordingly, held that "those who took affirmative steps directed toward the improper discharge of solvent wastes-for instance , by manufacturing a system designed to dispose of wastes improperly or by instructing users of its products to dispose of wastes improperly-may be liable under [ ( Wat. Code § 13304, subd (a) ) ] but those who merely placed solvents in the stream of commerce without warning adequately of the dangers of improper disposal are not liable under that section of the [code]." ( Modesto I , supra , 119 Cal.App.4th at p. 43, 13 Cal.Rptr.3d 865, emphasis added.) We issued a writ of mandate directing the superior court to reconsider defendants' motion for summary adjudication of the Polanco Act and negligence per se claims "in accordance with the views expressed herein." ( Id . at p. 45, 13 Cal.Rptr.3d 865.)

4. Defendants' Interpretation of Modesto I is Rejected

On remand, the trial court reconsidered the matter, and issued a tentative decision denying the motion. The court explained: "[With regard to t]he solvent manufacturer defendants, the Court of Appeal instructed that simply placing this material in the stream of commerce without warnings would not *148be enough, ... but if there were affirmative steps taken by the solvent manufacturers directed towards improper discharges, then that would be enough." The trial court found that plaintiffs had produced sufficient evidence to raise fact questions as to whether the manufacturers would be responsible parties under the Polanco Act.⁸ Seeking to convince the court otherwise, defendants argued that *777the Modesto I opinion addressed "only the first question in the chain of causation that has to be established;" the plaintiffs must also prove, they contended, that the instructions were received by the dry cleaners, that the dry cleaners acted in response to those instructions, and that their actions in response to the instructions were actions that caused the contamination. It was defendants' position that because Modesto I required an "affirmative step[ ] ... directed toward improper discharge," this was a "special circumstance ... where the Court of Appeal did not accept the idea that substantial factor itself was sufficient for liability under nuisance or under the Polanco Act." Hence, defendants argued, instead of the standard substantial factor causation test, the plaintiffs must prove the instruction gets to the dry cleaner, "[t]he dry cleaner looks at it, acts on it and causes the contamination."

The trial court pointed out, however, that the question before it was not whether it could make findings on causation, but only whether there were sufficient facts that would "rationally allow a juror" to make such findings. Rejecting defendants' arguments, the trial court affirmed its tentative ruling, concluding that, under the standard articulated in Modesto I , a reasonable juror could find a violation of the Polanco Act based on the evidence before the court.⁹

*149Defendants renewed their Modesto I causation theory at closing argument in the first Polanco Act trial (Phase II). Relying on the word "cause" in the Water Code, defendants argued: "Causation is built right into this. So in order to find liability under that Water Code section, not only do you have to start with an instruction that's improper, you have to get it to the site, somebody has to follow it, a release has to occur as a result, and that release has to cause the contamination that's at issue." Defendants insisted that this was precisely what Modesto I required.

The trial court again rejected defendants' interpretation of Modesto I . The judge identified the "critical inquiry" as "whether the defendants' [PCE manufacturers'] actions, taken as a whole, 'created or assisted in the creation of the nuisance.' " As that court explained, "[t]he appellate Court cited the giving of 'instructions' as an example of something that may 'assist in the creation' of a nuisance, but did not suggest that this is the only conduct that would qualify." The court noted that Modesto I identified one limitation on the otherwise broad liability for nuisance viz., that the law of nuisance is not intended *778to serve as a surrogate for ordinary products liability. It then reasoned: If " 'defendants who fail to warn of the dangers of improper disposal of hazardous materials but give no guidance or instructions pertaining to that disposal [pose] a more difficult question' [citing Modesto I ]," then "[d]efendants who do give 'guidance or instruction' do not pose a difficult question-they are covered by the Act."

The trial court then quoted the language by which it would be guided. " '[T]hose who took affirmative steps directed toward the improper discharge of solvent wastes-for instance by manufacturing a system designed to dispose of wastes improperly or by instructing users of its products to dispose of wastes improperly-may be liable under the statute, but those who merely placed solvents in the stream of commerce without warning adequately of the dangers of improper disposal are not liable under that section of the [Water Code].' City of Modesto, 119 Cal.App.4th at 43 [13 Cal.Rptr.3d 865] (emphasis added)." Applying that standard, the court recited its findings: "The manufacturer defendants in this case did more than simply place their PCE products into the stream of commerce without adequate warnings. Because their PCE products were fungible, the manufacturers competed in the marketplace by touting their expertise, professionalism, and individualized services. Their customers were relatively high volume businesses that used substantial amounts of PCE on a daily basis, and the manufacturers encouraged these businesses to rely on the manufacturers' advice, instructions and expertise. The manufacturers published newsletters to their customers, provided technical literature to their customers, and trained sales personnel to promote reliance by the customers on the manufacturers' expertise. [¶] .... [¶] The evidence included numerous examples of manufacturer instructions, advice, and guidance to customers to discharge separator water, which the manufacturers knew to contain PCE, *150into sewers, as well as to release waste PCE onto the ground. These kinds of recommendations were repeatedly made and reinforced by the manufacturers over the course of many years and led to the recommended PCE waste handling and disposal practices being generally followed by the customers. The effect of the manufacturers' recommendations, considered both individually and in combination, was that risky waste handling and disposal practices became the norm among customers." The court took note of the fact that a similar course of conduct had been found by the jury in Phase I to support liability for the creation of a nuisance at one of the dry cleaner sites.

The court found, in addition, that at the time the manufacturers were holding themselves out as experts, they knew that PCE could cause groundwater contamination; that Street employees visited Modesto Steam over the course of two decades and during those visits poured PCE from "titration tests" into the drain; and that the manufacturers delayed too long in correcting their communications to their customers concerning the improper waste handling and disposal practices. "The record, taken as a whole," the court concluded, "establishes that the manufacturer defendants are 'responsible parties' under the Polanco Act ... because they 'created or assisted in the creation of a nuisance.' "

5. Defendants' Modesto I Causation Theory is Accepted

The trial court's rejection of defendants' causation theory in the first Polanco Act trial did not deter them from making the *779same argument in the second Polanco Act trial, presided over by a different judge. Describing themselves as "remote manufacturers," defendants again characterized the Modesto I opinion as setting forth a "specific instruction test," which necessitated a showing of "an affirmative recommendation or instruction by a defendant to a dry cleaner that in turn was relied upon [ ] by the dry cleaner to discharge material to the environment which in turn caused the contamination which occurred here." "It's not only whether affirmative instructions existed during that timeframe[,] ... it's whether some dry cleaner came into court and said I read that instruction, and I relied on that instruction, and then whether an expert said that conduct caused the contamination we're talking about today."

The City, for its part, urged the court not to focus on the manufacturers' disposal instructions in isolation but to look at the totality of defendants' conduct over decades, in addition to the provision of instructions, to determine whether all of their activities in the marketplace, including a failure to take corrective action, could be found to have assisted in creating the PCE pollution, citing as an example, the findings of the judge in the Phase II trial. The City argued that the correct inquiry is "what did defendants do in *151addition to placing PCE into the stream of commerce without adequate warnings that would make this case more than simply a failure to warn case, and therefore subject defendants to Polanco Act [ ] liability?"

In the end, the trial court agreed with defendants. Although the court accurately summarized the reasoning and holding of Modesto I , it interpolated into that opinion the proof requirements which defendants had championed. Noting that the Polanco Act implicitly requires proof that "defendant's conduct caused the discharge in question," the court observed that Modesto I did not spell out the "kind of factual showing a plaintiff must make in order to prove that a solvent manufacturer's improper disposal instruction(s) caused a release of PCE. (Italics added.)" It nevertheless decided that specific proof requirements "follow" from Modesto I 's holding: "First, the user must have received the instruction(s). Second, the user must have disposed of hazardous waste in a manner consistent with the instruction(s). Third, the contamination for which cost recovery is sought under the Act must have been caused by a release mechanism contained in the instruction(s) and employed by the user[,] ... since the release could have been caused by an operator acting independently of a defendant's instruction(s) or by some other event unrelated to the actions of a defendant. Finally, with respect to the release of PCE, the user must have purchased the defendant's solvent, since the manufacturers in this case only provided PCE-related information to their customers and end users."

The trial court specifically rejected the City's argument that the court should consider whether other conduct-including whether defendants touted their expertise and individualized services with respect to PCE and its disposal; whether defendants promoted reliance by their customers on defendants' expertise; and whether, once defendants knew that PCE could contaminate groundwater, they failed to send out corrective information about disposal practices-to determine whether defendants were responsible parties. The court concluded that Modesto I neither requires nor even suggests that these factors are relevant. The court specifically rejected consideration of "whether Defendants' actions, *780taken as a whole, created or assisted in the creation of a nuisance." As we explain, this interpretation of Modesto I was incorrect.

6. Analysis

a. Modesto I did not address causation

To begin with, the trial court's focus on the provision of disposal instructions to the exclusion of all other potentially relevant factors is not supported by the holding of Modesto I . As the judge in Phase II observed, Modesto I *152held that "those who merely placed solvents in the stream of commerce without warning adequately of the dangers of improper disposal are not liable," but that those who take "affirmative steps" toward the improper discharge of solvents can be liable. Modesto I did not preclude consideration of any other factors that might be relevant to the question of whether such a defendant "assisted in the creation of [a] nuisance." ( Modesto I, supra, 119 Cal.App.4th at p. 38, 13 Cal.Rptr.3d 865.)

More to the point, Modesto I simply did not address the issue of how causation must be proven. Defendants argue that Modesto I requires proof of a special "chain of causation," demonstrating by direct evidence that a specific disposal instruction was received, read and acted on by a specific dry cleaner, which act caused contamination at a specific site. In support, they cite and emphasize phrases in the Modesto I opinion indicating that those whose involvement in discharges is "remote and passive" or who have "no active involvement" in activity that is "directly connected with the disposal of solvents" cannot be liable under the Polanco Act. But these phrases merely explain how we construed the term "responsible party" in the statute; they say nothing about proof of causation. As the trial court correctly observed, the Modesto I opinion did not address the quantum or nature of proof required to prove the causation element of liability. In short, nothing in the opinion either states or implies that any unusual or special causation test would apply.¹⁰

b. The substantial factor test of causation

Defendants also contend, however, that the substantial factor causation test requires this same heightened standard of proof. Defendants begin by reciting some basic legal principles, such as, that liability cannot be premised on a mere possibility of causation, nor on probabilities that are, at best, evenly balanced, nor on speculation *781or conjecture, citing Saelzler v. Advanced Group *153400 (2001) 25 Cal.4th 763, 775-776, 107 Cal.Rptr.2d 617, 23 P.3d 1143 and Merrill v. Navegar, Inc. (2001) 26 Cal.4th 465, 490, 110 Cal.Rptr.2d 370, 28 P.3d 116. No one disputes these principles. But this case is not akin to Saelzler , where plaintiff could not prove that additional security guards in a 300-unit, 28-building complex would have prevented a criminal assault ( Saelzler, supra, 25 Cal.4th at p. 777, 107 Cal.Rptr.2d 617, 23 P.3d 1143 ), or Merrill , where plaintiff provided no evidence either "direct or circumstantial" that the promotion and marketing of a firearm by a manufacturer bore any causal relation to the purchase and use of that firearm by an individual to kill various victims. ( Merrill, supra, 26 Cal.4th at p. 489, 110 Cal.Rptr.2d 370, 28 P.3d 116.) Defendants nevertheless argue that the principles enunciated in Saelzler and Merrill require, in this case, direct evidence linking a given instruction to a specific dry cleaner who would testify that he read and followed it, thence to expert testimony proving the dry cleaner's act was the mechanism of contamination. Otherwise, they argue, the evidence supports only the possibility that a defendant contributed to the contamination in a particular location. For example, defendants contend that even the conformance of a dry cleaner's disposal practices to a defendant's instructions is not sufficient to prove causation because "[s]uch happenstance does nothing more than allow for the 'mere possibility' that the defendant's instruction had something to do with the drycleaner's conduct." We do not agree that the substantial factor test of causation requires the kind of incontrovertible linkage proposed by defendants.

"Although a finding of causation may not be based on mere speculation or conjecture, such finding may be predicated on reasonable inferences drawn from circumstantial evidence." ( Smith v. Lockheed Propulsion Co. (1967) 247 Cal.App.2d 774, 780, 56 Cal.Rptr. 128.) Direct proof of each link in a chain of causation is not required. "[C]ircumstantial evidence of sufficient substantiality" from which reasonable inferences can be drawn will support a finding of causation in fact. ( Ibid. ) "Causation may in many instances be inferred from evidence that does not itself constitute direct evidence of reliance on an individual basis." ( State ex rel. Wilson v. Superior Court (2014) 227 Cal.App.4th 579, 608, 174 Cal.Rptr.3d 317 ( Wilson ).) "Just as factors such as the magnitude and temporal proximity of the unlawful conduct might evidence or negate the existence of fraud, so too might many of the same factors influence the extent to which an inference of causation is appropriate." ( Id. at p. 605, 174 Cal.Rptr.3d 317.)¹¹

Defendants cite Viner v. Sweet (2003) 30 Cal.4th 1232, 135 Cal.Rptr.2d 629, 70 P.3d 1046 ( Viner ) for the propositions that the substantial factor test subsumes the traditional requirement of "but for" causation, and that there can be no liability if any link in the chain of causation is missing. These statements are accurate, but do not answer the question posed-whether defendants' "specific instruction test" is the only means of proving that defendant's conduct was a substantial factor in causing the PCE pollution. In *154Viner the trial court ruled that the usual test of causation for attorney malpractice-that the plaintiff's outcome would have been better but for the attorney's negligence-did not apply to malpractice in the performance of transactional work. ( Viner , supra , 30 Cal.4th at p. 1238, 135 Cal.Rptr.2d 629, 70 P.3d 1046.) *782The court of appeal affirmed, holding that "a plaintiff suing an attorney for transactional malpractice need not show that the harm would not have occurred in the absence of the attorney's negligence." ( Id . at p. 1240, 135 Cal.Rptr.2d 629, 70 P.3d 1046.) The Supreme Court disagreed. The court of appeal effectively removed causation from the liability matrix, and that was error.

In Viner the high court found no reason to except transactional malpractice from the traditional "but for" standard applied in malpractice cases. ( Id . at pp. 1240-1242, 135 Cal.Rptr.2d 629, 70 P.3d 1046.) It went on, however, to caution that "the plaintiff need not prove causation with absolute certainty. Rather, the plaintiff need only ' "introduce evidence which affords a reasonable basis for the conclusion that it is more likely than not that the conduct of the defendant was a cause in fact of the result." ' [Citation.]" ( Id . at p. 1243, 135 Cal.Rptr.2d 629, 70 P.3d 1046.) And, of course, "the plaintiff may use circumstantial evidence to satisfy his or her burden." ( Id . at p. 1242, 135 Cal.Rptr.2d 629, 70 P.3d 1046.)

So, for example, in Wilson , plaintiff brought a qui tam action against a drug company (BMS) alleging it provided lavish gifts to physicians and members of formulary committees to induce them to prescribe BMS's drugs. Plaintiff alleged that "as a result of kickbacks BMS provided to them" the physicians and formulary committees selected BMS's drugs, and submitted insurance claims therefor, in violation of a statute prohibiting the employment of persons to procure patients to obtain insurance benefits. ( Wilson , supra , 227 Cal.App.4th at p. 587, 174 Cal.Rptr.3d 317.) The trial court granted summary adjudication on the following question: Assuming BMS provided an "item or service" of value to a doctor, and the doctor thereafter prescribed BMS's drugs, is the statute violated "absent proof that the item or service caused the prescription." ( Id . at p. 589, 174 Cal.Rptr.3d 317.) The trial court ruled that it was "not enough to prove that the unlawful conduct was a substantial factor resulting in the prescription." ( Id. at p. 590, 174 Cal.Rptr.3d 317.)

The court of appeal reversed. It concluded the trial court erred when it ruled that plaintiffs must establish not only that BMS's conduct was a substantial factor resulting in the prescriptions, but also that it was essential to the result, i.e., "that if the prescription would have been written even without BMS's unlawful inducement, the unlawful conduct cannot be found to have caused the prescription and claim." ( Wilson , supra , 227 Cal.App.4th at p. 607, 174 Cal.Rptr.3d 317.) The appellate court observed that the trial court's standard would make proof of an unlawful claim almost impossible to establish. ( Ibid. ) The court also noted that but-for causation is not required where there may be, as in that case, independent concurrent causes. ( Ibid. )

*155Similarly, in Stevens v. Parke, Davis & Co. (1973) 9 Cal.3d 51, 107 Cal.Rptr. 45, 507 P.2d 653 ( Stevens ), plaintiffs' family member died after her physician prescribed a drug manufactured by Parke, Davis, due to a condition induced by the drug. In addition to suing the prescribing physician, plaintiffs successfully sued Parke, Davis for "overpromoting" the drug, and "watering down" the warnings concerning the drug's link to the fatal condition. In the supreme court, the drug company argued its overpromotion and watered-down warnings were not proven to be the cause of death because the prescribing doctor admitted he was aware that the drug had been linked to the condition *783and that its "prolonged administration carried some danger of fatality." ( Id. at pp. 66-67, & fn. 15, 107 Cal.Rptr. 45, 507 P.2d 653.) The doctor had also testified that he obtained that information from articles in medical journals and from discussions with fellow physicians, and he could not remember specific instances in which he received any information, promotional or otherwise, directly from Parke, Davis, although he had received visits from drug salesmen and read journals which contained advertisements for the drug. ( Id. at p. 68, fn. 16, 107 Cal.Rptr. 45, 507 P.2d 653.)

The court affirmed the jury's verdict. It concluded there was "adequate circumstantial evidence in the record...to support a reasonable inference by the jury that [the physician] was induced to prescribe the drug for [the decedent] because of Parke, Davis' overpromotion. Like many others of the profession, he had been exposed to the promotional tactics employed by Parke, Davis. It is reasonable to assume that the company's efforts consciously or subconsciously influenced him."¹² ( Stevens , supra , 9 Cal.3d at p. 68, 107 Cal.Rptr. 45, 507 P.2d 653 ; see also, Toole v. Richardson-Merrell, Inc. (1967) 251 Cal.App.2d 689, 705, 60 Cal.Rptr. 398 [trial court correctly rejected defendant's proposed special causation jury instruction requiring a direct causal link between plaintiff's injury and defendant's violation of law].) These authorities support the conclusion here that direct proof of every link in the chain of causation, on a site-by-site and instruction-by-instruction basis, is not required. Liability can be proven by sufficient circumstantial evidence that would allow a reasonable fact-finder to find that all of defendants' conduct-to include affirmative steps toward the discharge of toxic wastes-was a contributing factor to the pollution. As has been noted, "the State Water Board has concluded that even a relatively minor contribution to a discharge may support a finding of responsibility." ( Modesto I, supra, 119 Cal.App.4th at p. 41, 13 Cal.Rptr.3d 865.)

*156As in Stevens , the absence of proof that a dry cleaner consciously followed an improper disposal instruction-or even a statement that a dry cleaner did not rely on a disposal instruction-does not ipso facto break the chain of causation. To be sure, it was the dry cleaners and not the manufacturers who discharged the PCE. But " ' "[t]he substantial factor standard is a relatively broad one, requiring only that the contribution of the individual cause be more than negligible or theoretical." [Citation.] Thus, "a force which plays only an 'infinitesimal' or 'theoretical' part in bringing about injury, damage, or loss is not a substantial factor" (citation), but a very minor force that does cause harm is a substantial factor' (citation)." ( Bettencourt v. Hennessy Industries, Inc. (2012) 205 Cal.App.4th 1103, 1123, 141 Cal.Rptr.3d 167.)

Defendants also contend that product identification is, itself, always required for liability because one manufacturer may not *784be held liable for the injury resulting from another manufacturer's product. ( O'Neil v. Crane Co. (2012) 53 Cal.4th 335, 135 Cal.Rptr.3d 288, 266 P.3d 987 ( O'Neil ).) There are exceptions, however, and one is where "the defendant's own product contributed substantially to the harm...or because the defendant participated substantially in creating a harmful combined use of the products." ( Id. at p. 362, 135 Cal.Rptr.3d 288, 266 P.3d 987.) Defendants contend, more precisely, it was not proven that during the years their product was provably sold at any Phase IV site, an improper disposal instruction was disseminated. Connecting those dots is one way to prove causation; but other evidence, both direct and circumstantial, can also demonstrate that the manufacturers' instructions and other conduct were substantial factors in causing the pollution. Thus, in keeping with common law nuisance principles ( Modesto I, supra, 119 Cal.App.4th at p. 38, 13 Cal.Rptr.3d 865 ), "[i]n the ordinary case, the liability ... arises because one person's acts set in motion a force or chain of events resulting in the invasion. The acts may be a direct and immediate cause of the invasion, ...or they may be an indirect cause of the invasion...." ( Rest.2d Torts, § 824, com. 6, p. 116.)

Finally, defendants argue that permitting abatement liability to be imposed on manufacturers without proof of each causal link-"that a specific affirmative instruction was received, read, and acted on by a particular drycleaner, and then resulted in contamination at a given site"-would be adverse to public policy, citing Ferguson v. Lieff, Cabraser, Heimann & Bernstein (2003) 30 Cal.4th 1037, 135 Cal.Rptr.2d 46, 69 P.3d 965 ( Ferguson ) and O'Neil , supra , 53 Cal.4th 335, 135 Cal.Rptr.3d 288, 266 P.3d 987. Neither these authorities nor any such policy issues control here.

In Ferguson , our high court held that allowing plaintiffs in a malpractice action to recover, as compensatory damages, the punitive damages their *157attorneys failed to recover from the tortfeasor, would violate public policy in various ways. Primarily, it would contravene the entire purpose of punitive damages which is not to compensate the plaintiff but to punish the tortfeasor and deter him or her from similar egregious conduct. ( Ferguson , supra , 30 Cal.4th at p. 1046, 135 Cal.Rptr.2d 46, 69 P.3d 965.) In addition, it would exact too great a social cost because it would hinder the courts' ability to manage and resolve mass tort claims, increase the cost of malpractice insurance, and discourage the use of mandatory non-opt-out class actions for punitive damages. ( Id. at pp. 1049-1050, 135 Cal.Rptr.2d 46, 69 P.3d 965.) None of these considerations apply here.

In O'Neil , the plaintiff had not been exposed to any asbestos-containing product manufactured or sold by the defendant, but only to the products of others used to replace, many years later, the packing and gaskets in defendant's pumps and valves. The court declined to impose liability because a manufacturer could not " 'reasonably be expected to foresee the risk of latent disease arising from products supplied by others that may be used with the manufacturer's product years or decades after the product leaves the manufacturer's control.' " ( O'Neil , supra , 53 Cal.4th at pp. 364-365, 135 Cal.Rptr.3d 288, 266 P.3d 987.) The court further reasoned that imposing a duty on such an attenuated basis would not be consistent with public policy, not only because the manufacturer's conduct was lacking in moral blame, but also because such a duty would not be likely to prevent future harm, and it was "doubtful that manufacturers could insure against the 'unknowable risks and hazards' lurking in every product that could possibly be used with or in the manufacturer's product."

*785( Id . at p. 365, 135 Cal.Rptr.3d 288, 266 P.3d 987.) Again, no such concerns are at play here.

Defendants cite to the principle that " 'proximate cause "is ordinarily concerned, not with the fact of causation, but with the various considerations of policy that limit an actor's responsibility for the consequences of his conduct." ' [Citation.]" ( Ferguson , supra , 30 Cal.4th at p. 1045, 135 Cal.Rptr.2d 46, 69 P.3d 965.) But defendants have not applied that principle to the facts of this case nor to their causation argument. Defendants have asserted no policy reasons for restricting their liability here; instead they merely repeat their claim that Modesto I intended to impose this "restriction[ ] on nuisance liability." But we have rejected defendants' interpretation of Modesto I , and it provides no support for defendants' policy argument. If anything, the social costs of limiting the responsibility of chemical manufacturers under defendants' formulation would fall far too heavily on the victims of the pollution by setting an almost insurmountable standard for proving liability. (Cf. Wilson, supra, 227 Cal.App.4th at p. 607, 174 Cal.Rptr.3d 317.)

The measure of proof for causation should be no different here than that applied in any other similar action. As noted previously, the plaintiff need only " ' "introduce evidence which affords a reasonable basis for the conclusion that it is more likely than not that the conduct of the defendant was a *158cause in fact of the result." ' [Citation.]" ( Viner , supra , 30 Cal.4th at p. 1243, 135 Cal.Rptr.2d 629, 70 P.3d 1046 ; accord, People v. ConAgra Grocery Products Co. (2017) 17 Cal.App.5th 51.)

c. CERCLA Liability Standards Do Not Apply

The City advanced below and advances here an alternative argument to support causation. It contends that the Polanco Act incorporated by reference CERCLA's "scope and standard of liability" (§ 33459.4, subd. (c)) and therefore, under CERCLA case law, the defendants bear the burden of proving that they did not cause the release that triggered the response costs, citing United States v. Alcan Aluminum Corp. (2d Cir. 1993) 990 F.2d 711, 721 ( Alcan ). The trial court correctly rejected that argument. As was explained in Redevelopment Agency v. Salvation Army (2002) 103 Cal.App.4th 755, 127 Cal.Rptr.2d 30, section 33459.4 of the Polanco Act incorporated CERCLA's liability standards, "to wit, strict liability regardless of knowledge or intent [citation], joint and several liability [citation], and retroactive liability [citation]." ( Id. at p. 766, 127 Cal.Rptr.2d 30.) But these liability standards do not negate or supplant the requirement to prove causation. As we have explained, the Polanco Act incorporates by reference the HSAA/CERCLA categories as one set of "responsible parties" but plaintiffs do not contend defendant manufacturers fall within any of those categories. The shifting burden of proof found in those schemes only applies in that context. (See Orange County Water District v. Alcoa Global Fasteners, Inc. (2017) 12 Cal.App.5th 252, 306-309, 219 Cal.Rptr.3d 474 [CERCLA and HSAA govern liability for specific categories of defendants; as to those defendants once a plaintiff has proven that a release has occurred at a site and has triggered response costs, the burden of disproving the provenance of the discharge shifts to the defendant]; see also Asarco LLC v. NL Industries, Inc. (E.D. Mo. 2015) 106 F.Supp.3d 1015, 1031 [if a defendant fits into one of the four categories of responsible parties " 'it is enough that response costs resulted from "a" release or threatened release-not necessarily the defendant's release or threatened *786release' "].) The HSAA and CERCLA liability standards do not apply here.

7. Conclusion

Defendants argued and the trial court concluded that Modesto I set up a special proof standard under the Polanco Act, viz, that causation can be demonstrated only by proof that a specific disposal instruction was received by a dry cleaner who read, relied on and followed the instruction and by that act caused the contamination, to the exclusion of any other evidence that might be relevant to prove defendants assisted in creating the pollution. This is not a fair reading of the opinion. Modesto I did not set limits on how causation was to be proven; it articulated a liability standard that required, as a threshold matter, that the provider of the product be something more than a mute, passive supplier. "Those [manufacturers] who took affirmative steps directed toward the improper discharge of solvent wastes-for instance, ... by instructing users of its products to dispose of wastes improperly-may be liable under [that statute]" but those who merely placed solvents in the stream of commerce without warning about the dangers of improper disposal are not liable. ( Modesto I , supra , 119 Cal.App.4th at p. 43, 13 Cal.Rptr.3d 865.)

*159That formulation does not change the causation analysis, which is, considering the evidence as a whole, is it more likely than not that defendants' improper instructions, and any other relevant conduct, were a substantial factor in causing the pollution. Because the trial court imposed a far more stringent proof of causation requirement in the Phase IV trial, we shall vacate that portion of the judgment.

Under the substantial factor test there may or may not be sufficient evidence to support liability for the Phase IV sites but that question has not been put to us, and will have to be resolved on remand.¹³

B.-H.^††

V. DISPOSITION

The order denying reconsideration of the summary adjudication of the nuisance claims is vacated and the matter is remanded with directions to deny the motions for summary adjudication on the nuisance claims and to conduct further proceedings.

The punitive damages award against Dow in the Phase I proceeding is reversed.

*160The order granting the Motion for Directed Verdict Re Property Damage in the Phase III proceeding is vacated, except as to any of those sites for which nonsuit was granted on other grounds, and the matter is remanded for further proceedings.

The decision and order denying relief in the Phase IV proceeding is reversed and the matter is remanded for further proceedings.

The further proceedings ordered herein shall not include R.R. Street and Company.

The amended judgment is vacated as void.

The prevailing party determinations with respect to Phases III and IV are *787vacated as a result of our decision to vacate the Phase III directed verdict and the Phase IV decision and order.

The order allocating settlement credits is vacated.

In all other respects we affirm the judgment entered on November 15, 2011.

All parties shall bear their own costs on appeal except for those costs comprised of attorney's fees to which a party may be entitled pursuant to statute, which shall be determined in the first instance in the trial court.

We concur:

Ruvolo, P.J.

Streeter, J.

Certified for Partial Publication.^*

* Pursuant to California Rules of Court, rules 8.1105(b) and 8.1110, Sections I, II, III.B. and Section IV.A. (except subsection 1. ) of this opinion are certified for publication.

1 All unattributed statutory cites are to the Health and Safety Code.

** See footnote *, ante .

*** See footnote *, ante .

† See footnote *, ante .

6 The federal statute provides that four categories of persons or entities are, in essence, strictly liable for clean-up costs resulting from hazardous substances: (1) those who own or operate the facility; (2) those who previously owned or operated the facility when the hazardous substances were released; (3) those who arranged for disposal or treatment, or for transportation for disposal or treatment, of the hazardous substances found at a treatment facility; and (4) those who accepted hazardous substances for transport, disposal or treatment at a site chosen by that person, that results in a release or threatened release. (42 U.S.C. § 9607(a).)

7 The "assistance and advice" alleged to have been received from the chemical suppliers is not described in the opinion and, as we describe below, it does not play any part in the court's liability analysis or its holding.

8 For context, here are examples of the instructions and guidance that was provided by Dow for disposal of PCE and PCE waste. Material Safety Data Sheets (MSDS's) were given to dry cleaners through distributors when they purchased PCE. In 1971 the MSDS's instructed users to flush large spills to the ground and to mop up small spills and bury them. Throughout the rest of the 1970's, the MSDS's instructed that if there were "small leaks" of PCE the user should "[m]op up, wipe up or soak up immediately. Remove to out of doors." The "disposal method" prescribed for PCE was to send it to a reclaimer, but "[i]n some cases it can be transported to an area where it can be placed on the ground and allowed to evaporate safely." Similar instructions regarding "small leaks" were provided in MSDS's issued in the 1980's. In that period, however, the MSDS's "strongly discouraged" dumping unused PCE "into sewers, on the ground or into any body of water." Dow's "Dry Cleaning: A Basic Handbook DOWPER Solvents for the '80s," providing "information and instruction" about solvent and dry cleaning equipment, was among the literature distributed not just to customers, but also at trade shows, seminars and by mail. In the discussion on equipment, these handbooks indicate that contact water from the water separator is "drawn off and discarded."

9 In their Phase I trial brief (on the strict liability and negligence claims) defendants made a similar causation argument. They argued the manufacturer had to be directly linked to a release of its product, thence to the pollution. Defendants repeated that argument in a motion for nonsuit, in closing argument to the jury, and yet again in motions for judgment notwithstanding the verdict. Neither the jury nor the judge were persuaded.

10 In connection with this issue, we asked the parties to address at oral argument the nuisance jury instruction relied upon by the court in the Phase II Polanco Act trial, because the record suggested that defendants had agreed this was the applicable standard. As pertinent here, the instruction states, (1) the City claims "it suffered harm because one or more of the defendants created a public nuisance;" (2) to establish that claim the City must prove (inter alia) that a defendant "affirmatively instructed users to dispose of wastes improperly resulting in a condition that obstructed the free use of city property ," and that "the defendant's conduct was a substantial factor in causing harm to the City." [Emphasis added.] At oral argument, defendants contended the italicized part of the instruction separately required a heightened degree of proof of a direct link between a disposal instruction and the identified contamination. We disagree. The language on its face requires no such direct linkage, but only proof that defendants' disposal instructions resulted in obstructing free use of property and that defendants conduct was a substantial factor in creating that harm. We also note, in passing, that it does not appear any such heightened degree of proof was used when the jury applied this instruction in Phase I.

11 In their petition for rehearing, defendants disclaim that they ever "insisted ... that causation could be proven only by direct evidence." But defendants did insist-and argued that the trial court insisted-on proof that dry cleaners read, relied upon and followed defendants' instructions. "As Judge Goldsmith made clear, the evidence must show that the reason the drycleaner [disposed of the PCE waste consistent with defendants' instructions] was that she or he read and acted on [ ] the instruction[s]." (Italics added.) Defendants do not explain how those facts could be proven other than by direct evidence.

12 The court also alluded to expert testimony that advertising and promotion of the drug "played a role" in inducing physicians to prescribe the drug when it was not sound practice to do so. But the court appeared to conclude that even in the absence of that testimony, the jury could find that the doctor's negligent prescription of the drug was a "foreseeable consequence of the extensive advertising and promotional campaign." (Stevens , supra , 9 Cal.3d at p. 69, 107 Cal.Rptr. 45, 507 P.2d 653.)

13 It is reasonable to expect that on remand liability could be found if the correct causation standard is utilized. This is because: (1) the vast bulk of the evidence pertaining to the Polanco Act bench trials was presented during the preceding jury trials (Phases I and III); (2) application of the traditional causation test resulted in a finding of liability under the Polanco Act in Phase II based on the evidence in Phase I; and (3) according to defendants , the "same facts" were adduced in Phases I and III.

†† See footnote *, ante .