Impact of Continuous HMA Air Voids Measurements on PWL for Construction Quality Evaluation

Nondestructive technologies, such as Density Profiling Systems (DPS), enable continuous in-field measurement of air voids and density across the entire project, overcoming the limitations of spot checks from a few quality assurance (QA) cores. Current agency practices rely on sample statistics—sample mean \(\left(\overline{\mathbf{x}}\right)\) and standard deviation (s) to estimate population parameters (μ, σ) and calculate Percent Within Limits (PWL). While the sample mean is an unbiased estimator of the population mean, its variability decreases only as sample sizes increase. However, the sample standard deviation (s) is inherently noisier, with a slight downward bias in small samples that can underestimate variability and inflate PWL. This statistical uncertainty propagates into PWL and Pay Factor (PF) estimates, where small or sparse samples often misrepresent true lot quality, increasing the risk of overpayment if PWL is overestimated or underestimated. By contrast, continuous measurements capture the true spatial variability across the entire pavement, reducing reliance on inference and producing more accurate, stable PWL and PF estimates. The results presented in this paper show that continuous air-void (AV) measurements can substantially improve PWL accuracy. As a result, this will reduce the agency's risk of accepting poor construction quality. For agencies, such an approach would minimize sampling risk, better align payments with actual material quality, and strengthen fairness and transparency in acceptance decisions.