delivered the opinion of the court:
Plaintiff, Ivan Tepper, appeals the judgment of the circuit court granted in favor of the defendants, the County of Lake, the Public Service Committee of the County of Lake, and the Lake County Public Works Department (hereinafter collectively referred to as the Waterworks). Plaintiff contends the trial court erred in granting a directed finding in plaintiff’s declaratory judgment action. Plaintiff contends that he presented a prima facie case that the Waterworks sent him a bill for water service based on an inaccurate water meter. At the heart of this cause is whether plaintiff proved he did not use 126,200 gallons of water which was registered on the meter.
Plaintiff and his wife, Wendy, were nine-year residents of Grays-lake in Lake County. They received a bill for 126,200 gallons of water for the period between June 13, 1989, through September 17, 1989, for which they were charged $365.98. Plaintiff attempted to resolve the dispute with the Waterworks, but when he did not pay, the Waterworks threatened to impose a lien upon the property. Plaintiff filed an action in the circuit court in which he sought a variety of relief, including an injunction, a declaratory judgment and administrative review, for which he named the individual members of the board as defendants.
At this point, we note that municipally owned utilities are expressly excluded from the Public Utilities Act (Act) (Ill. Rev. Stat. 1991, ch. 1112/3, par. 3—105(1)). Thus, the provisions of that Act and the regulations of the Illinois Commerce Commission (see, e.g., 83 Ill. Adm. Code §600.310 (1991) (standards of accuracy)) are of limited value in analyzing the principles involving the provision of utility service. (See Village of Niles v. City of Chicago (1990), *82201 Ill. App. 3d 651, 663-64.) At most, the statutes evidence a legitimate public concern for the accuracy of all meters. (See Ill. Rev. Stat. 1991, ch. 1112/3, par. 8—301; Peoples Gas, Light & Coke Co. v. Illinois Commerce Comm’n (1988), 175 Ill. App. 3d 39, 49.) The Waterworks alleged that its actions were not subject to administrative review. The only cause of action litigated before the trial court involved the declaratory judgment action.
That this cause was litigated as a declaratory judgment is significant because plaintiff carried the burden of proof. Stripped of the details in procedure, this action involves a contract to supply water. (See Rosborough v. City of Moline (1961), 30 Ill. App. 2d 167, 179.) The consumer has a contractual obligation to pay for water provided by the municipality. (Brooks v. Village of Wilmette (1979), 72 Ill. App. 3d 753, 756.) The Waterworks alleges that it delivered goods, namely, 126,200 gallons of water, and wants its bill paid. Ordinarily, a seller of goods would commence a suit to collect an action for the price. The Waterworks, however, had the power to place a lien on the property and shut off the connection, and it declined to initiate the suit. It thus shifted the onus to plaintiff to file suit to resolve the lien. The plaintiff in a declaratory judgment action has the burden of proof. (Board of Trade v. Dow Jones & Co. (1982), 108 Ill. App. 3d 681, 686-88, aff'd (1983), 98 Ill. 2d 109, 116.) Similarly, if plaintiff had paid the bill under protest with a reservation of his rights (see Ill. Rev. Stat. 1989, ch. 26, par. 1—207 (Uniform Commercial Code provides for performance of contract under reservation of rights); see also Community Convalescent Center of Naperville, Inc. v. First Interstate Mortgage Co. (1989), 181 Ill. App. 3d 996, 999 (suit to recover payment made under protest)), plaintiff would still have to initiate suit to recover his overpayment.
In a contract action, the Waterworks would have to prove that it delivered 126,200 gallons of water. The Waterworks cannot produce any direct testimony of a witness with personal knowledge of the delivery of the water. Instead, it would produce the readings of the meter as evidence of the usage and a witness to vouch for the reliability of the meter pursuant to the scientific evidence or business records tests. (See 2 J. Wigmore, Evidence §§665, 665a (Chadbourn rev. ed. 1979); cf. People v. Orth (1988), 124 Ill. 2d 326, 340 (State must lay foundation for breath-alcohol meter).) In the declaratory judgment action, the consumer has the burden of presenting some evidence that the reading of the meter or the meter itself was not reliable.
*83Pursuant to the Waterworks’ motion, the trial court entered judgment in favor of defendants at the conclusion of plaintiff’s case. The trial court found that plaintiff failed his burden of proof and that the loss of water could have been explained by a stuck toilet over a three-week vacation or some other explanation.
Plaintiff contends that the trial court improperly granted defendants’ motion at this point of the trial. When considering such a motion in nonjury causes, the trial court must first consider whether a plaintiff has made a prima facie case by presenting some evidence concerning each element of the cause of action. If he has, the trial court then must consider all of the evidence, including any in favor of the defendant, determine the credibility of the witnesses, draw reasonable inferences from the testimony and weigh the evidence. (Kokinis v. Kotrich (1980), 81 Ill. 2d 151, 154-55.) This weighing process may result in the negation of some of the evidence necessary to prove the plaintiff’s prima facie case, in which event the trial court must enter judgment in the defendant’s favor. (81 Ill. 2d at 155.) A reviewing court will not reverse the decision of the trial court to grant a directed finding unless it is against the manifest weight of the evidence. Dyduch v. Crystal Green Corp. (1991), 221 Ill. App. 3d 474, 477.
Before we analyze the evidence presented in plaintiff’s cause, we find it helpful to view the bare numbers in perspective, which neither party attempts to do. Plaintiff contends that he was billed for 60,000 gallons more than his normal use. Over a 20-day period, such as when plaintiff was on vacation, this usage is equivalent to 3,000 gallons per day. This usage is approximately 120 gallons per hour for 24 hours, or a rate of 2 gallons per minute. Because toilets are common household utilities, a fact finder would be able to utilize his common sense and experience in the affairs of life to determine whether a flow rate of two gallons per minute would be reasonable when viewed in light of all of the other evidence presented.
Plaintiff attempted to prove his case by offering his and his wife’s testimony that they did not accept the water and did not use the water. While they could not testify to the amount of water they did use, they testified that they used less water than they normally did in a three-month summer. They supplied the facts which supported their opinion, including their personal recollections that they did not use the water. They offered the water bills for other quarters, which they summarized as follows:
*84“Old New
Dates Reading Reading Used Bill
[gallons] [$]
12/19/88-03/20/89 161,700 177,700 16,000 46.60
03/20/89-06/13/89 177,000 231,500 53,800 160.66
06/13/89-09/14/89 231,500 357,700 126,200 365.98
09/14/89-12/18/89 357,700 10,200 16,320 47.33
12/18/89-03/16/90 10,200 26,600 16,400 47.56
03/16/90-06/13/90 26,600 46,300 19,700 57.13
06/13/90-09/18/90 46,300 118,100 71,800 208.22
09/18/90-12/17/90 118,100 140,500 22,400 64.96
12/17/90-03/13/91 140,500 161,900 21,400 62.06”
They testified that their water usage was less than that of previous years because only in that year did they depart on a three-week vacation, during which they used no water. They also had personal knowledge that the weather was rainy in the summer of 1989; thus, they did not irrigate the garden or add water to the swimming pool. They supported their allegations with a certified copy of the daily precipitation records of the National Climatic Data Center for the months of June, July, August and September 1989. Plaintiff testified that he drained the pool in May, prior to the billing period, and found no leaks from his inspection done by noting the algae line and by looking for hydrostatic pressure. They testified plaintiff filled the pool in May. No one was present or had access to the house during the vacation. The house had two outdoor spigots which were not locked. They did not own a humidifier, water softener, icemaker or any other device which used water automatically. They observed or noticed no leaks or thefts of water. Wendy Tepper’s daughter, Morgan, also lived at the house, but she did not testify.
Plaintiff also offered the testimony of Michael Faircloth, who was the meter division supervisor of the Waterworks. After plaintiff complained about his bill, Faircloth went to plaintiff’s house on October 17, 1989. Faircloth inspected all of plaintiff’s fixtures and found no fixture to be leaking. He attached a flow finder to the meter and turned off each fixture in the house. The flow finder registered that no water was moving through the pipes; thus, there were no leaks in plaintiff’s house’s plumbing. Faircloth repeated these tests on October 24 and found no leaks. Faircloth removed the meter and replaced it with a new one. He took the old meter to the Waterworks office for testing.
*85While the trial court did not admit any evidence during plaintiff’s case to demonstrate that the meter was accurate, plaintiff introduced evidence which may have impeached the validity of the Waterworks’ accuracy tests, the results of which were not yet before the court. Faircloth testified that the Waterworks tested a water meter by attaching it to a bench testing device. The examiner turns a valve, runs water through the meter at three different rates, and calculates the accuracy of the meter at each rate. When Faircloth performed his test, he failed to record the volume running through the meter at each rate and recorded only his accuracy calculation. He did not compare the meter reading to the output in a volume container. He had no knowledge of whether the valve measuring the water flowing into the meter had itself been tested for accuracy. Plaintiff offered the testimony of Martin Galantha, who was the superintendent of the Waterworks. He testified that a valid test could not be performed on a meter without the use of a volume container. Galantha was not aware of whether the test valve had ever been tested for accuracy.
These facts do not attack the reliability of the meter, but merely impeach prospectively the results of the Waterworks’ accuracy tests. The results of the accuracy tests had not been admitted in plaintiff’s case. The readings of the meter were admitted into evidence, and the parties do not dispute that the meter registered 126,200 gallons for the period. The only dispute was whether the meter was registering accurately.
Plaintiff offered scant evidence to prove the meter was inaccurate other than the circumstantial evidence that he and his wife did not remember using the water. When Faircloth removed the meter from plaintiff’s house, he tagged it and gave it to William Bartlett at the Waterworks. Before he died, Bartlett performed a test on the meter and put it on a shelf. Faircloth’s accuracy test results differed slightly from Bartlett’s. The parties stipulated that Bartlett recorded the meter reading at the conclusion of his test on October 24, 1989, as 0363950, and that when Faircloth performed his test on June 3, 1991, the starting reading was 0364160. The meter thus showed a 210-gallon difference between the two tests. Faircloth testified that he had no knowledge of anyone using the meter in the year that the meter was in his office.
Faircloth also made statements concerning the accuracy of the class of meters generally. He stated that the meters do not overre-gister but sometimes underregister the amount of water. If the internal gears get out of line, they cease to record rather than over-*86record. However, Wendy Tepper testified that Faircloth told her that unexplained losses of water occurred all over Lake County. Faircloth admitted at trial that within Lake County the Waterworks experienced an average of six occurrences per year of unexplained water losses. Galantha testified that the Waterworks experienced three to five occurrences per year of disputed readings. The Waterworks maintained 15,000 accounts for water service.
Since the trial court directed a judgment in favor of the Waterworks, we must determine if plaintiff established a prima facie case or if some of the elements necessary to the prima facie case were negated by the weighing of the evidence. (Dyduch, 221 Ill. App. 3d at 477.) The standard for granting the motion is not, as the trial judge suggested, one of the preponderance of the evidence. At this stage in the proceedings there is no contrary evidence against which to weigh the plaintiff’s evidence. The plaintiff’s evidence is uncontradicted and must be accepted if credible. The function of the trial court at this stage is to determine whether the elements of the prima facie case were negated. See Kokinis, 81 Ill. 2d at 155.
To obtain guidance, we refer to a similar action where the evidentiary battle is between man and machine. (People v. Orth (1988), 124 Ill. 2d 326, 340-41.) In Orth, our supreme court determined that the defendant in a driving-under-the-influence-of-alcohol proceeding has the burden of proof in a civil proceeding to rescind the summary suspension of his driver’s license. The driver is the party seeking judicial relief. (124 Ill. 2d at 337.) The driver must present a prima facie case that the result of the breath-alcohol meter did not accurately register his alcohol concentration. (124 Ill. 2d at 340.) Once the motorist presents this evidence, the burden shifts to the State to move to admit the results of the meter into evidence and to lay the necessary foundation. (124 Ill. 2d at 340.) The supreme court left the trial court to determine whether the prima facie case was established. (124 Ill. 2d at 341.) Where the motorist argues that the meter results were unreliable, his evidence may consist of any circumstance which tends to cast doubt on the meter’s accuracy and may include credible testimony by the motorist that he was not in fact under the influence of alcohol. Only if the trial court finds such testimony credible will the burden shift to the State to lay a proper foundation for the admission of the test results. 124 Ill. 2d at 341.
Orth is analogous because the human party bears the burden of proof by a preponderance of evidence in a civil proceeding, has to prove a negative, and must disprove the reliability of a machine not *87yet in evidence. While a water meter is not as prone to inaccurate calculations as a breath-alcohol meter (see Orth, 124 Ill. 2d at 336), the foundational requirements of introducing a water meter will not be as stringent as those of the breath-alcohol meter (see Orth, 124 Ill. 2d at 340).
In the cause before the court, plaintiff had to prove a negative. He had to prove that he did not take the water. He has presented some circumstantial evidence to establish a prima facie case. Under Orth, an informal opinion, if credible, showing a contrary result of the meter reading is enough to cast the meter reading into doubt. He testified that he did not use an unusual amount of water. The meter reading was not merely 10% or 20% in excess of his normal bill, but 200% or 300% above any amount he had ever taken. He presented some evidence that he had eliminated other possible explanations. His testimony that he did not use the water for the garden or the pool was supported by objective records. The trial court specifically stated that plaintiff’s and his wife’s testimony was credible. Thus, there was some credible evidence which passed the evidentiary threshold and which cast the accuracy of the meter into doubt. The burden thus shifted to the Waterworks to lay the foundation and to introduce its evidence of the accuracy of the meter. Thus, the trial court erred by entering judgment in favor of the defendants at the conclusion of plaintiff’s case.
Pursuant to Supreme Court Rule 366(b)(3)(iii) (134 Ill. 2d R. 366(b)(3)(iii)), the cause must be remanded to proceed as though the motion had been denied by the trial court or waived.
For the above reasons, the order of the circuit court of Lake County is reversed, and the cause is remanded for further proceedings.
Reversed and remanded.
WOODWARD and DOYLE, JJ., concur.