Some of Gelman's own supply wells showed double-digit levels of 1,4-dioxane as early as 1989:
This image is from pages 79-83 in a 1993 printout of Gelman's sampling. (The units shown are in mg/l, so 0.012 mg/l = 12 ppb.) More pages are available in the Gelman archives at the various repository locations such as the Ann Arbor District Library and Scio Township Hall.
Production Wells #1 & #2 (GSSW1 & GSSW2) are Gelman water supply wells used prior to being hooked up to city water but do not show on the company's plume maps. According to well logs obtained from the state, GSSW1 was 157 feet deep and GSSW2 was 152 feet deep. There are only a few other wells around that depth at the time...
What's curious is that Gelman claimed that an extensive clay layer protected deeper aquifers from contamination around that depth. Here is on of the earliest cross-section diagrams the company provided (in 1995), but the original didn't show its own supply wells that reach 152-159 feet deep and were already contaminated (added by SRSW in yellow 03/13/2017):
The company also failed to sample a key deep well it had east of Wagner Road from Sept 1993 until Feb 2001 even though that well, MW-30d, was 9 ppb in 1993, 3 times the 3 ppb cleanup standard then). When it was finally sampled again in 2001, the levels had risen to 67 ppb. After the company changed hands in 1997, that well's earlier sampling pre-2001 was left off the new sampling database and has been left off ever since.
Even today, there are few wells on the original Gelman site that are screened near the same elevation as their original supply wells: