OPINION
NIERENGARTEN, Judge.Appellant Jennifer Rae Haagenson’s driver’s license was revoked pursuant to the implied consent law for failing the In-toxilyzer test. She petitioned for judicial review, challenging the reliability and accuracy of the test. The trial court sustained the revocation and she appeals.
FACTS
Deputy Sheriff Switajewski arrested appellant Jennifer Rae Haagenson for driving while under the influence. Switajewski, a certified Intoxilyzer operator, administered an Intoxilyzer test to Haagenson at the sheriff’s office in Walker, Minnesota. He testified that Haagenson was under his observation for at least 15 to 20 minutes prior to the administration of the test, and in fact had been under his observation since the arrest. She had nothing to eat or drink after the arrest and he did not notice that she put anything in her mouth which could have changed the results of the test or adversely affected the test in any way.
Switajewski testified that the simulator solution used in the Intoxilyzer test is obtained from the Bureau of Criminal Apprehension, which certifies the alcohol concentration at .11. He explained that the simulator solution for the Intoxilyzer is changed every month or at the completion of an Intoxilyzer log. About seventeen tests were run on the simulator solution prior to Haagenson’s test. Switajewski testified that the amount of alcohol in the simulator solution is depleted when that number of tests is run.
Switajewski detailed the steps for the Intoxilyzer test which were recorded on the test record. A diagnostic test indicated the instrument was working properly; the instrument was at the proper operating temperature. Four air blank tests were run, all of which were at .000, the anticipated result. The first breath sample yielded a reading of .224 and a replicate reading of .230. The simulator solution was then tested. The first reading was .099, and the replicate reading was .100. The first read*435ing was below the accepted standard by .001. The officer testified nothing indicated to him that the instrument was working improperly. The second subject breath sample gave a reading of .226, with a replicate reading of .226. The breath correlation was 100%, and the reported value was .22.
The officer testified that he did not receive any information which would lead him to believe the test results erroneously indicated an alcohol concentration in excess of .10, nor that the .099 simulator solution was the result of anything other than the depletion of the simulator solution.
On cross-examination, he testified that he was trained by the BCA that the simulator solution has to read between .10 to .12 for a valid test, and that in this case one of the readings was .099. He also testified that during the observation period, he was with Haagenson, but did not take any notes. He testified that the chief dispatcher has maintenance responsibility for the Intoxilyzer. Switajewski did not review the maintenance records for the In-toxilyzer and did not know whether it was repaired during the time period they were discussing. He said he thought the log would indicate whether the Intoxilyzer was “down,” but said there was no such indication on the log.
The officer also testified on redirect examination that he did not hear any burping or regurgitating. To the best of his knowledge, the Intoxilyzer remained in service from May 26 to the time of trial. He also testified that three days after Haagenson’s test, another test yielded a simulator solution test result below .100.
The trial court made findings that the test results were accurate, that the fact that the simulator solution read .099 did not adversely affect the test results, and that the reliability of the test was established. It sustained the revocation, and Haagenson appeals.
ISSUE
Did the test accurately and reliably indicate that appellant’s alcohol concentration was .10 or more?
ANALYSIS
Haagenson challenges the trial court’s determination that the test was reliable. She argues that the low simulator solution reading renders the test inaccurate, and contends that the observation period was insufficient.
Haagenson argues there is some question as to whether she was observed for 15 minutes, because Officer Switajew-ski did not take notes on the observation period. However, he testified that she was under his observation for at least 15 to 20 minutes prior to the test, and he did not notice her put anything into her mouth which could have affected the test result. There was no evidence indicating the observation period was faulty or that she put any objects in her mouth. The observation period was clearly adequate.
Haagenson also challenges the reliability and accuracy of the test results because of the low simulator solution reading. This court has addressed the issue of a low simulator solution in several cases:
The Commissioner has the burden of establishing that the administration of the Intoxilyzer test conforms to the procedure necessary to ensure reliability; the opponent must then suggest reasons why the test was untrustworthy. State v. Dille, 258 N.W.2d 565, 567, 568 (Minn.1977); Tate v. Commissioner of Public Safety, 356 N.W.2d 766, 767-68 (Minn.Ct.App.1984). A low simulator solution reading does not automatically render a test unreliable. Johnson v. Commissioner of Public Safety, 374 N.W.2d 577, 579 (Minn.Ct.App.1985). There must be some indication that the low reading would unduly exaggerate the test results. Noren v. Commissioner of Public Safety, 363 N.W.2d 315, 318 (Minn.Ct.App.1985).
Kadrlik v. Commissioner of Public Safety, 388 N.W.2d 8, 10-11 (Minn.Ct.App.1986).
The officer testified that the simulator solution was changed every month *436or at the end of the log. Seventeen tests were run using the same simulator solution prior to Haagenson’s test. The officer testified that the amount of alcohol is depleted when that number of tests is run. All other diagnostic tests- for the machine were within acceptable limits, the breath correlation was 100%, and the reported value of Haagenson’s breath was .22. The officer made no observations that would make him suspect the results erroneously indicated an alcohol concentration in excess of .10 and received no information which would make him believe the .099 simulator solution was the result of anything other than the depletion of the alcohol in the simulator solution.
Haagenson argues that the Intoxilyzer log does not support the officer’s testimony that the solution became depleted as tests were run. However, while there are several peaks, there is also a downward trend, particularly with the last five tests, supporting the officer’s testimony. Based on the facts in this case, the trial court properly found that the Commissioner proved the reliability of the test by a preponderance of the evidence.
DECISION
The order of the trial court sustaining the revocation is affirmed.
Affirmed.