Attached are plots of the flow rates through heads [1,4] for the last 12 hours and the last hour.  Apart
from the initial restart period, the flow rates more or less settled down and looked reasonably clean.
However in the last hour they took a downturn back to rates similar to when the chiller was restarted
last night.  The crystal chiller flow rate appears constant at this point in time, at least within the
minute trend data.

    The crystal chiller flow rate has a resolution of about -/+ 0.1 lpm.  The observed decrease in the
flow rate to the heads is ~ -0.025 lpm.  So in retrospect, perhaps the observed decrease in the head flow
rates is due to a reduced flow from the crystal chiller.