Identification and solution of common problems

An estimate of the blank for filtration and storage on-board ship and shore laboratory analytical steps has been determined by passing replicate photo-oxidised seawater samples through the procedure. The mean overall blank was 6/tmol/L DOC with a coefficient of variation of 10 % (n=5), giving a detection limit (taken as 3cr of the blank) of 2 / mol DOOL. The blank of the analytical procedure alone (with no persulphate or blank additions) is about 2 umol/LDOC (Gershey et al., 1979). Replicate analyses of two stored seawaters gave mean DOC concentrations of 57timol/LDOC (coefficient of variation 6.0%, n=30) and 92 /imol/L DOC (coefficient of variation 2.4 %, n=30). 

This section describes problems that have been encountered using the analyser, and their remedies general problems associated with autoanalyser systems are covered in Chapter 10 of this volume. 

A greatly reduced lamp life (/.e., less than 100h) can be caused by differential cooling of the electrodes. The aluminium tubing supporting the silica shielding as shown in Fig. 16-3 is intended to prevent cooling of the lower end of the lamp. 

---