Capacitive Deionisation (CDI) as a desalination technique proves to be one of the most promising technologies owing to its simplicity and ease of manufacturability. The longevity of CDI depends on a number of factors, such as the ability of carbon to withstand Faradaic reactions to CO2. Furthermore, like other technologies such as (reverse osmosis) RO and (electrodialysis reversal) EDR, scale formation is known to affect its hydraulic efficiency. The slow mineral deposition of magnesium and calcium carbonate is responsible for such phenomena. The extent of the problem is one where the hydraulic efficiency of the device drops well below 10%, a state where the device is considered to be at the end-of-life as it is too difficult to pump water through it. In the preliminary study we examined the performance of an industrial scale (membrane capacitive deionisation) MCDI module that has been in operation intermittently for 5 years to establish familiarity with the performance of a used module. In the main study, we described a new technique for observing the chemical state within the spacer channel of a CDI cell, the part of the electrode that is prone to blockage. This new technique for the direct measurement of pH is based on colourimetric analysis of a novel synthetic formulation of a pH-sensitive dye dissolved in the electrolyte phase. The basis for this novel technique combined a simple optical microscope integrated with a microfluidic device. This microfluidic device containing a microreactor cell of 750 picolitres in size was fabricated to host the electrochemical reaction (electrode and electrolyte phases) demonstrated in this study that recreates the “CDI effect” of an industrial cell. The pH indicator formulations were synthesised from a 1:1:2 ratio of phenolphthalein, bromothymol blue and methyl red in an ethanol solution. The pH indicator, which we denote as the standard formulation (SF), had a density of 0.93 kg/L and pH of 2.74. This was the formulation that was subsequently used in the main part of the study. The dye was characterised by visible range spectroscopy and subsequent conversion to (hue saturation value) HSV colourmap coordinates using the 1976 CIE (International Commission on Illumination) chromaticity model. This enabled the development of a range of pH-Hue relationships using best approximated by sigmoid functions. In the analysis of these functions, we investigated the sensitivity of the pH-Hue relationship to variation in the SF formulation due to selective electrosorption of the dye. It was shown that the variation in functions produced was not significant and that the general pH-Hue relationship did not break down into a non-usable form. Application of the novel technique was initially performed on a membraneless cell configuration. The pH profile was investigated from a starting range of interest of pH 8 to 9 where scale formation occurs. In the second series of reactions, we demonstrated microreactor CDI configurations with varying degrees of membrane coverage from 12.5% to 100%. It was found that the pH variation in the spacer channel was minimal or non-existent across the full charge cycle from adsorption through to desorption. This indicates that obscuration of the carbon surface blocks the protons diffusing from the surface of the carbon where it is generated by way of the standard faradaic reaction in CDI or blocking co-ions from being expelled from the membrane that may result in speciation that affects the pH. We recommend longer in-situ observation type studies be performed to confirm the speciation chemistry present in the spacer channel typical of many situations where electrodes are reducing in performance by way of fouling and scale formation.