Table 3.

Personal exposures to 1,3-butadiene (µg/m³) at Refinery 1 and 2

Exposure group	n/N	% >LOD	µ X	95% CI	Max	µ Y	${\mathit{\sigma }}_Y^2$	${\mathit{\sigma }}_{bY}^2$ (%)	${\mathit{\sigma }}_{wY}^2$ (%)
Refinery 1
Refinery process technicians	68/132	44	5.4	3.1–9.5	975.8	−0.1	3.5	16	84
Outdoor process techniciansa	59/108	50	7.2	3.9–13.4	975.8	0.2	3.6	15	85
Process areab	40/71	44	3.6	1.9–6.9	79.5	−0.1	2.8	21	79
Harbour and tank parkb	19/37	62	22.4	9.4–53.3	975.8	0.8	4.7	6	94
Indoor process techniciansa	14/24	17	0.7	0.3–1.6	6.4	−1.2	1.8	21	79
Refinery 2
Outdoor process technicians	45/66	38	1.8	1.1–2.9	90.8	−0.4	2.0	31	69
Process area	35/51	33	1.4	0.8–2.4	15.9	−0.5	1.7	47	53
Tank park	10/15	53	4.2	1.0–18.2	90.8	−0.1	3.0	11	89
Loading of LPGc	4/13	100	15.6	7.4–33.1	31.3	2.2	1.2	0	100

n number of workers, N number of measurements, % > LOD percentage of samples above the limit of detection (LOD), µ _X arithmetic mean (AM), calculated as exp(µ _Y + ${\mathit{\sigma }}_Y^2$/2) where µ _Y and ${\mathit{\sigma }}_Y^2$ are the mean and the variance of the log-transformed observations, CI confidence interval for the AM, Max maximum level detected, µ _Y and ${\mathit{\sigma }}_Y^2$ mean and variance of the log-transformed observations, ${\mathit{\sigma }}_{bY}^2$ and ${\mathit{\sigma }}_{wY}^2$ between- and within-individual variance of the log-transformed observations

^aSignificant difference in geometric mean (GM) between outdoor and indoor process technicians, p < 0.05

^bSignificant difference in GM between outdoor process technicians working in the process area and the harbour and tank park, p < 0.05

^cRailroad terminal workers performing loading of liquified petroleum gas (LPG) on to railroad tank cars