delivered the opinion of the court:
On January 29, 1985, plaintiffs Donald D. Malone and Velva M. Malone sued defendant Ware Oil Company in the circuit court of McLean County. Plaintiffs’ complaint alleged contaminants released or deposited by Ware Oil Company (Wareco) from its automobile service station’s premises seeped through the ground to the basement area of plaintiffs’ nearby residence, which was located downhill from the Wareco station. Plaintiffs maintained the gasoline contaminated their basement by giving rise to an extremely foul smell. Plaintiffs maintained Wareco was guilty of the torts of private nuisance and negligence. After a jury trial was held, the court entered judgment on April 14, 1988, on a general jury verdict in favor of plaintiffs and against Wareco in the sum of $36,000.
Wareco has appealed, contending the evidence does not support the verdict and the court erred in rulings on instructions and Wareco’s objections to plaintiffs’ closing arguments. The parties do not dispute the rule stated in Patterson v. Peabody Coal Co. (1954), 3 Ill. App. 2d 311, 122 N.E.2d 48, that, in order to recover for a private nuisance, plaintiffs must- show (1) there was a substantial invasion of the use and enjoyment of their land; and (2) the invasion was either negligent or intentional and unreasonable. We conclude the evidence was insufficient to support a jury determination that any invasion of plaintiffs’ property by gasoline coming from Wareco’s property resulted from acts of commission or omission by Wareco which were either negligent or intentional and unreasonable. Thus, the verdict is not supported by either the nuisance or the negligence charges. We must reverse the judgment appealed and need not consider the other contentions of Wareco.
It is undisputed from the evidence that Velva Malone purchased a home at 409 North Livingston, in Bloomington, Illinois, in approximately 1955. She and Donald married in 1967, and they lived in the Bloomington residence from 1968 to the present. At all times relevant, Wareco owned and operated a gasoline service station on property located northeast and uphill from plaintiffs’ home. The surrounding neighborhood was zoned “light industrial” and included numerous industries.
The parties agree that in approximately 1976 Wareco replaced its above-ground upright storage tanks on the premises with underground tanks. Plaintiffs’ evidence indicated that about a year thereafter, Donald and Velva Malone began smelling an odor resembling gasoline in their home, and the odor persisted on and off since then. Velva testified the odor was strongest afterxheavy rains. She stated *732that when the odor was present, it started in the basement around the sump pump and rose into the kitchen and hallway. Donald testified he was not greatly bothered by the odor, but Velva testified the odor made her eyes, nose, and throat dry. A real estate appraiser, testifying on behalf of plaintiffs, estimated their property to be worth (1) $22,500 absent any odor problem; (2) $11,000 with the odor problem, assuming the problem could be cured at a reasonable cost; and (3) only $1,500, after demolition, if the odor problem could not be cured.
The evidence also undisputedly showed that on March 2, 1978, Donald Malone complained about the odor to the Bloomington fire department. The fire department investigated the complaint but reached no conclusion as a result of the investigation. The evidence indicates Donald Malone also made subsequent complaints to the fire department.
The fire department then drew samples from plaintiffs’ sump pump and from numerous tanks in the area and sent them to an Illinois State University chemist for analysis. The chemist hired by the fire department testified that, based upon his analysis, there was no question the sump pump contaminant was something similar to a typical gasoline. However, the chemist also stated he was unable to “match” the sump pump samples with any of the gasoline samples taken from the various tanks in the neighborhood. He further stated that, of the 15 samples he examined, six of them matched up better than the remaining nine samples. He said he was unable to differentiate among the samples which more nearly resembled the sump pump sample. A sample from the Wareco tanks was among those six samples.
The fire department further requested Wareco to test its tanks to determine whether they were leaking. The tests revealed Wareco’s six tanks did “leak,” but those leaks totaled .127 gallons per hour and were well within the city ordinance leakage limit of .505 gallons per hour. Therefore, the tanks were considered “tight.” The fire marshal stated the company which performed the tests may not have taken the water table into account in testing the tanks. Several witnesses indicated this omission could have flawed the test results one way or another because the water would determine the amount of pressure on the tanks, and, the greater the pressure, the less the leakage.
Following these tests, the fire department sent a letter to plaintiffs which indicated “Wareco was not the source of the odor” in the Malones’ house. (Emphasis added.) The fire marshal further stated at trial that no evidence had exclusively established defendant as the source of the odor on those times that the fire department had re*733sponded to plaintiffs’ complaints.
In 1983 and 1984, employees of the Illinois Environmental Protection Agency (EPA) installed four groundwater monitoring wells near the Wareeo service station. Independent consultants hired by defendant also sunk wells on and about defendant’s property. The EPA took soil borings from the ground as well.
The results of the EPA tests generally indicated that the well located north of and uphill from defendant showed no detectable level of gasoline in it, while wells located south of and downhill from the service station contained varying levels of gasoline concentration. A witness testified the test results from Wareco’s consultants “corroborated and corresponded very closely to the EPA results.” However, the actual test results from that firm were not introduced into evidence.
Plaintiffs also hired their own consulting firm to analyze the odor problem. A geologist from that firm reviewed the following documentation sent to him: (1) an Illinois EPA report which documented the monitoring wells and testing groundwater; (2) the test results of testing performed on tanks at the station site to determine whether Wareco’s tanks were “tight”; (3) Bloomington fire department records; and (4) the test results from monitoring wells located on the station site. Based upon his review of those records, he testified to an opinion “that the Wareeo station site [was] in fact the source of gasoline encountered in wells and in the sump of the Malone residence.” He could not identify the exact source of the contamination but concluded the concentrations of gasoline found on plaintiffs’ property were too high to be the result of day-to-day spillage by defendant.
Many witnesses acknowledged the remote possibility that the source of the contamination could have been a spill years ago and miles up-gradient from the service station. Based upon the evidence given as to the source of the gasoline, however, we conclude the evidence was sufficient to support a jury determination that gasoline from defendant’s service station had migrated underground to plaintiffs’ property and caused the contamination.
Defense witnesses testified concerning the care with which the Wareeo station was operated. A station manager explained the procedures used by defendant to monitor the amount of gasoline received and disbursed to customers. He said the transport driver gives them a bill of lading which contains the number of gallons received in the transport. The driver measures the tank with a stick both before and after the gasoline is delivered, and those figures are also entered on the bill of lading given to defendant. In addition, the manager mea*734sures the tank daily to determine if any of their product is missing. Those numbers are compared with the meters on the pumps or the electronic console. Several witnesses indicated the Bloomington station had not suffered any loss of its product during the time period in question.
A gasoline transport driver for defendant testified concerning the systems used by the station to fill its tanks. He said that prior to 1986, the system used to fill the tanks was not airtight. He stated that, in the filling process, a cupful of fuel commonly spilled out, and there was no way to prevent it. After 1986, however, an airtight system was installed, with no spills resulting. The station received three or four deliveries per week.
The station had its lines reworked in September 1984. A fire inspector present on the premises during that time said he observed only one joint with a very slight dampness and found no product which he could see that had leaked into the ground. The lines again tested “tight” at that time. The station then removed its six underground tanks in September 1986. At that time, it further discovered two additional aged skeleton tanks which it also removed.
A fire inspector monitored the removal of these tanks in 1986. He said he viewed only a slight amount of soil contamination around three of the tanks. Although he observed spill contamination around the large 8,300-gallon tank which necessitated removal of IV2 truck loads of contaminated soil, he said the amount of contamination was not large considering the size of the tank. Testimony indicated spill contamination could be a result of many actions or inactions, including overfilling of the tanks upon delivery of the fuel.
Clearly, the jury could properly have determined from the evidence that the contaminant causing the odors in plaintiffs’ residence was gasoline which had migrated from the Wareco station to plaintiffs’ property. We then confront the question of whether the jury could have determined the migration resulted from Wareco’s intentional and unreasonable or negligent conduct.
We examine first the question of intent as applied to the intentional and unreasonable factor of the tort of private nuisance. In considering this issue, the Patterson court looked to the Restatement of the Law on Torts, where the word “ ‘[ijntentional’ is further defined in paragraph 825 as including a knowledge that the invasion of another’s interest *** is resulting or is substantially certain to result.” (Patterson, 3 Ill. App. 2d at 316, 122 N.E.2d at 51; citing Restatement of Torts §825 (1934).) The present version of that test similarly defines the conduct of an actor causing a nuisance as intentional *735if the actor “(a) acts for the purpose of causing it, or (b) knows that it is resulting or is substantially certain to result from his conduct.” (Restatement (Second) of Torts §825, at 117 (1979).) The intent element of the tort of private nuisance is also defined as the creation or continuance of the condition giving rise to the invasion “with full knowledge that the harm to the plaintiff’s interests are occurring or are substantially certain to follow.” W. Keeton, Prosser & Keeton on Torts §87, at 625 (5th ed. 1984); see also Woods v. Khan (1981), 95 Ill. App. 3d 1087,1089, 420 N.E.2d 1028,1030.
Plaintiffs’ complaint contains no direct allegation Wareco acted intentionally, and the main thrust of its argument is not directed to that theory. We conclude the jury could not properly have based its verdict for plaintiffs on such a theory. For much of the time after plaintiffs began smelling gasoline in their home in 1978, great uncertainty existed as to the source of that gasoline. Despite the efforts of both Wareco and plaintiffs, the information that test wells showed gasoline to be migrating under the surface of the soil downhill from the Wareco station but not uphill from the station was not available to Wareco until a few months before suit was filed. Wareco had its tanks tested by a firm specializing in that work. Although the tests performed may have lacked accuracy because of failure to consider the water pressure in the area around the underground tanks, the report of the testing purported to show that the tanks were reasonably retaining the stored gasoline. The evidence did not show anything Wareco could have done to prevent the migration of gasoline to plaintiffs’ property other than to stop all operations and drain its tanks. Any determination by the jury that Wareco acted through its agents and servants with the intent required for the tort of nuisance would be contrary to the manifest weight of the evidence, and no contrary verdict could ever stand. Fitzpatrick v. City of Chicago (1986), 112 Ill. 2d 211, 492 N.E.2d 1292.
The complaint was also very general in its allegation of negligence which was asserted as the basis for either the nuisance or the negligence charges. No specific acts of commission or omission were alleged. To show negligence on Wareco’s part, plaintiffs rely on the evidence that (1) after suit was filed, Wareco’s underground tanks were removed and some contamination of surrounding soil was shown; (2) two abandoned underground tanks were found when the others were replaced; (3) the amount of tank leakage shown by the report of the independent firm; (4) the evidence of spillage upon the filling of the tanks; and (5) Wareco’s method of record keeping.
After the suit had been filed, Wareco did have its underground *736tanks replaced. The testimony concerning contamination indicated that the contamination around most of the tanks was minimal, but that V-k truckloads of soil had to be removed around the largest tank. Even in regard to that tank, however, the testimony indicated the contamination was not great considering the size of the tank. Moreover, the information concerning the contamination was available to Wareco only after the suit had been filed. Wareco is responsible in this suit only for conduct occurring up to the time of the commencement of the suit, as no supplemental complaint was filed. (See 61A Am. Jur. 2d Pleading §289, at 281 (1981).) This evidence does not constitute notice to Wareco which would bear upon the care it exercised at times pertinent.
The testimony indicated the skeleton tanks were empty, and no indication is given that they caused the contamination migrating to plaintiffs’ property. The report of the independent firm did show a total leakage of some .127 gallons per hour from the underground tanks. However, no evidence indicated that amount was unreasonable, and the amount was shown to be only a fraction of that permitted by a municipal ordinance. While evidence at trial indicated the tests may have been inaccurate because of the failure of the independent firm to consider the outside water pressure upon the tanks, any negligence in that respect was not chargeable to Wareco. It was given assurance its tanks were in good condition. Moreover, no evidence was presented as to what the tests would have shown as to leakage had proper consideration been given to the water pressure. The jury could not properly infer that the results would have shown unreasonable leakage.
The evidence did indicate that a cupful of gasoline per delivery was spilled in filling the tanks, and this occurred approximately three times a week. This amount of spillage was obviously so small as to be insignificant. Plaintiffs’ complaint in regard to Wareco’s record keeping was that Wareco did not use a computer to determine the amount of fuel remaining in the tank instead of using a dip stick. Again, the jury cannot infer from the evidence that ordinary care required the use of the computer.
Citing Haynes v. Coca Cola Bottling Co. (1976), 39 Ill. App. 3d 39, 350 N.E.2d 20, and Woodrick v. Smith Gas Service, Inc. (1967), 87 Ill. App. 2d 88, 230 N.E.2d 508, plaintiffs point out that, if the inferences arising from the circumstances “are reasonable in the light of the jurors’ ordinary sensibility,” a tort case may be proved by circumstantial evidence alone. They then assert that the evidence of some leakage and spillage at the service station property and the evidence that the gasoline flowed to plaintiffs’ premises is sufficient, of itself, to in*737fer the invasion of plaintiffs’ home was proximately caused by Ware-co’s negligence. Plaintiffs’ theory is very similar to the doctrine of res ipsa loquitur. (See Kolakowski v. Voris (1980), 83 Ill. 2d 388, 415 N.E.2d 397.) It is inapplicable here, though, because substantial quantities of gasoline can, in the ordinary course of affairs, escape from premises where a service station is being operated and travel underground to nearby properties downhill from the station without negligence on the part of the owner or operator of the station.
A jury determination here that Wareco was guilty of the negligence alleged to support either the nuisance charge or the simple negligence charge would also have been contrary to the manifest weight of the evidence, and a verdict based thereon could never stand. (Fitzpatrick, 112 Ill. 2d 211, 492 N.E.2d 1292.) Thus, the verdict cannot stand. We need not consider the question of whether the invasion of plaintiffs’ property was sufficiently substantial to meet the requirements of Patterson v. Peabody Coal Co. (1954), 3 Ill. App. 2d 311, 122 N.E.2d 48. Nor, as we have indicated, do we need to determine whether error occurred during the trial.
We reverse.
Reversed.
McCULLOUGH, P.J., and SPITZ, J., concur.