Kill slipped and fell on a wet, smooth metal rim of a utility cover in a downtown Seattle sidewalk. She sued the City, alleging that the utility cover rim was unreasonably dangerous. Kill relied on expert testimony about the slip-resistance of the rim measured by a tribometer. However, the tribometer calibrations fell outside the confidence interval specified by the manufacturer. As a result, the trial court excluded the slip-resistance testimony as unreliable and unhelpful to the jury under ER 702. The trial court held that the expert's remaining testimony was essentially that metal is slippery when wet, which is common knowledge and did not create a genuine issue of fact. The trial court accordingly dismissed on the City's motion for summary judgment. We affirm.
On November 13, 2009, Twyla Kill was walking along a sidewalk in downtown Seattle around 1:30 p.m. She recalled that it was raining at the time.
The utility cover—also known as a handhole—is diamond plated and is surrounded by a smooth, two-inch metal rim. The City owns the utility cover and it is in the City right of way. The City did not know when the utility cover was installed, but it was likely "in 1989 (when the Bank Centre building was completed) or before." There were no prior complaints about the specific utility cover or other similar utility cover rims.
Kill hired Joellen Gill as an expert to conduct tests and testify about the condition of the metal rim. To do so, Gill used an English XL Variable Incidence Tribometer, an instrument that measures the coefficient of friction, or "slip-resistance" of a surface. Gill used the tribometer to conduct two slip-resistance tests of the utility cover rim: the first in February 2011 and the second in June 2013.
Different standards for tribometer validation and calibration were in effect at the time of each of Gill's field tests. Prior to September 2006, the American Society for Testing and Material (ASTM) F1679 standard provided instructions for how to use a tribometer. However, the ASTM withdrew the F1679 standard in September 2006 and did not adopt a new standard for five years.
In March 2011, the current standard for tribometer validation and calibration, ASTM F2508, went into effect. ASTM F2508 states that "[v]alidation shall be performed by walkway tribometer suppliers or independent testing facilit[ies]." It defines "supplier" as "any individual, agent, company, manufacturer, or organization responsible for the walkway tribometer prior to receipt by the user." Thus, an individual user cannot validate a tribometer.
Under ASTM F2508, a tribometer must satisfy two criteria to be validated: (1) it must rank the coefficient of friction for each of four reference surface tiles in the correct order; and (2) it must produce statistically significant results, using the mean and standard deviation, for all adjacently ranked surface tiles. ASTM F2508 at § 9. If the tribometer does not satisfy these criteria, then it fails validation.
ASTM F2508 also requires individual users to perform calibration of their tribometer to ensure valid test results.
Therefore, ASTM F2508 ensures tribometer reliability with two safeguards. First, manufacturers or independent testers must validate the tribometer. Second, individual users must calibrate their tribometer to ensure their measurements fall within the confidence interval set forth in the validation report. If the tribometer fails either calibration or validation, then it fails to comply with ASTM F2508.
On February 24, 2011, just before the adoption of ASTM F2508, Gill tested the metal rim of the utility cover. She explained that the manufacturer calibrated her tribometer in January 2011. Gill found that the rim's coefficient of friction when wet was 0.35 (± 0.02). Gill opined that 0.35 is very slippery and not reasonably safe for pedestrian use. She stated that the "generally accepted standard is that 0.5 is the established
The coefficient of friction scale ranges from 0.0 to 1.0. In 2011, the City adopted a 0.5 coefficient of friction standard for new utility covers. Prior to that, the City did not have a standard in place.
On June 2, 2013, Gill conducted a second test of the metal rim. In an effort to recreate the wet conditions when Kill slipped, Gill poured water on the rim. Gill found that the coefficient of friction was 0.21 (± 0.02). Explaining the different results (0.35 versus 0.21), Kill stated that Gill explained "there must be some surface contaminant of some kind that was on the rim as tested that resulted in the lower figure the second time around."
The day before Gill's second test, she calibrated her tribometer and created what she called a "Report of ASTM F2508 Validation" of the English XL tribometer. She used the four reference tiles—granite, porcelain, vinyl, and ceramic—sent from ASTM. Her test results showed the coefficient of friction for each tile as: 0.0700 for granite, 0.1013 for porcelain, 0.1727 for vinyl, and 0.8505 for ceramic. However, the manufacturer's validation report for Gill's tribometer specifies the 95th percentile confidence intervals as: 0.078-0.082 for granite, 0.132-0.137 for porcelain, 0.173-0.180 for vinyl, and 0.605-0.616 for ceramic. None of Gill's calibration results fall within these intervals.
On the parties' cross motions for summary judgment, the trial court held Gill's methodology for slip-resistance testing to be unreliable and therefore unhelpful to the jury under ER 702. The court believed Gill was qualified as an expert. However, the court explained that Gill's tribometer was not properly calibrated when she tested the utility cover rim. Furthermore, the court reasoned, Gill's two tests of the rim produced different results. It concluded that Gill did not provide an adequate explanation for this difference. "Either Ms. Gill's tribometer is inherently unreliable or the way she used it was inherently unreliable." And, the court held that Gill failed to account for how the presence or absence of surface contaminants may have affected her test results.
As a result, the court excluded Gill's test results and her opinion that the rim was unreasonably slippery. Without Gill's testimony,
The court therefore held that Kill failed to demonstrate an issue of fact. It granted the City's motion for summary judgment and dismissed all claims against the City with prejudice.
After the trial court's ruling, Gill sent her tribometer to the manufacturer (Excel Tribometers LLC) for additional testing. Using Gill's tribometer and its own reference tiles, Excel produced the following calibration results: 0.066 for granite, 0.116 for porcelain, 0.165 for vinyl, and 0.576 for ceramic. Like Gill's original calibration, these results are outside the 95th percentile confidence interval in the validation report.
Excel also conducted a calibration test using its own tribometer and Gill's reference tiles. These results were: 0.080 for granite, 0.100 for porcelain, 0.149 for vinyl, and 0.641 for ceramic. Except for the granite tile, these results are also outside the 95th percentile confidence interval and all substantially different than Gill's calibration results for the same tiles. Excel also determined that the coefficient of friction for Gill's ceramic reference tile varied widely from quadrant to quadrant, ranging from 0.645 in the southeast quadrant to 0.840 in the southwest quadrant. Based its testing, Excel asserted an additional margin of error of ± 0.03 for slip resistance values equal to or less than 0.50 and ± 0.05 for slip resistance values greater than 0.50. Therefore, Excel believed that Gill's tribometer satisfied the ASTM F2508 calibration requirements.
In a motion for reconsideration, Kill stated:
Kill explained that ASTM F2508 assumes the tiles' surfaces are not variable, even though different tiles are used for validation and calibration. She argued that the difference between the validation report and Gill's calibration was attributable to this variation in tiles.
Kill also requested a
The trial court denied Kill's request for a
Gill's Expert Testimony
Kill argues that the trial court erred in excluding Gill's testimony, because Gill's opinions and methodology regarding slip-resistance testing are helpful and reliable. Kill makes several related arguments. She asserts that any minor margin of error or variance in Gill's test results goes to the weight and credibility of Gill's testimony, not its admissibility. She likewise argues that the trial court improperly overlooked the additional margin of error asserted by the tribometer manufacturer. Kill also contends that, because Gill's individual tribometer was successfully validated, there was no need for the additional ASTM F2508 calibration protocol.
Standard of Review
We review an order granting summary judgment de novo.
We generally review evidentiary rulings for abuse of discretion.
The trial court must exclude expert testimony involving scientific evidence that does not satisfy both
Evidence that is admissible under
Gill's Methodology and Test Results Were Unreliable
Kill is correct that "[w]hen a scientific theory has protocols for assuring reliability, an expert's errors in applying proper procedures go to the weight, not the admissibility, of the evidence."
Similarly, variance in test conditions and original conditions do not necessarily preclude admissibility, as variations may merely go to the weight of the evidence.
Here, Gill's test results varied so significantly as to render them unreliable. Her first test showed a coefficient of friction of 0.35 (± 0.02). Her second test showed 0.21 (± 0.02). This is a difference of 0.14, almost 15 percent of the entire coefficient of friction scale, which ranges from 0.0 to 1.0. This is not a minor variance.
Furthermore, there is no documentation in the record of the rim's condition on the day Kill slipped, other than it was wet. Gill poured water on the rim for her second test, but was able only to speculate that some type of surface contaminant must have caused the variation in test results. Without more evidence, Gill could not establish substantial similarity between her two tests, nor between the test conditions and the condition of the rim when Kill fell. As Gill's test results and explanation demonstrate, surface contamination can have a dramatic impact on the slipperiness of the metal rim.
Furthermore, Gill's tribometer was not properly calibrated under the ASTM F2508 standard at the time of her second field test. The following table shows the various calibration results:
The confidence interval represents a margin of uncertainty. Every measurement is uncertain, in that no instrument is infinitely precise or accurate.
The table demonstrates that Gill's tribometer measurements for each of the four surface reference tiles fell outside the manufacturer's 95th percentile confidence interval.
Both Kill and the City put forth evidence that ASTM F2508 is the industry standard. The standard specifies that if calibration results for each tile do not fall within the specified confidence interval, the tribometer fails calibration. ASTM F2508 § 13. A tribometer that fails calibration does not comply with ASTM F2508 and does not produce valid test results.
And, lastly, as the trial court held, Gill's statement that a 0.5 coefficient of friction is an absolute threshold for safety would mislead the jury. The City provided copious evidence that tribometers are effective to measure only relative slipperiness, not absolute slipperiness. Different tribometers, both across models and within models, can give significantly different readings for the same surface. Thus, a tribometer can accurately determine whether one surface is more or less slippery than other surfaces measured by that same tribometer, but it cannot give an objective measurement of a surface.
For all these reasons, we conclude that Gill's expert testimony on slip-resistance is unreliable and misleading. It would therefore be unhelpful to the jury under ER 702. As such, we hold that the trial court properly excluded the evidence.
The Trial Court Did Not Need to Hold a Frye Hearing
In the alternative, Kill argues that the trial court should have held a
Gill stated in her declaration that the English XL tribometer is generally accepted in the scientific community. The manufacturer stated that this particular tribometer "conforms to all requirements of ASTM F2508-11 Standard Practice for Validation and Calibration of Walkway Tribometers Using Reference Surfaces." The City submitted several scholarly articles explaining that slip-resistance measurements are tribometer specific. "The introduction of [ASTM F2508] . . . has produced a method which allows validation of each type of tribometer and the values generated during testing." This evidence demonstrates that the ASTM F2508 is generally accepted in the relevant scientific community and capable of producing reliable results. Without it, the literature indicates that tribometer results are not reliable. The fact that Gill could not conform her tribometer measurements to ASTM F2508 does not mean that ASTM F2508 is unreliable.
Genuine Issue of Material Fact
We must still decide whether there is a genuine issue of material fact in Kill's remaining evidence. Municipalities have a duty to exercise reasonable care to keep their sidewalks in a condition that is reasonably safe for ordinary travel.
The fact that the plaintiff slipped and fell does not, by itself, mean that there is an unreasonably dangerous condition.
Taking all reasonable inferences in Kill's favor, there is evidence that the utility cover rim was (1) wet from rain, (2) smooth metal, and (3) two inches wide. Kill also stated in her declaration that she was wearing work boots at the time of her fall. It is common knowledge that smooth, wet metal is slippery.
Kill nevertheless argues that the City did not comply with its own standards and practices, which she asserts is admissible to show the rim was not reasonably safe. In determining whether the defendant acted with reasonable care, the trier of fact may be informed of the standard industry practice. RESTATEMENT (SECOND) OF TORTS § 295A (1965);
Kill is correct that since at least 2003, the City has required all new utility cover frames to have a nonskid surface.
The City's current requirement of nonskid utility cover rims, by itself, does not establish that smooth metal rims are unreasonably dangerous.
VERELLEN and BECKER, JJ. concurs.
These cases are distinguishable from Kill's case.