Bacteria-based self-healing concrete is an innovative self-healing materials approach, whereby bacteria embedded in concrete can form a crack healing mineral precipitate. Structures made from self-healing concrete promise longer service lives, with associated economic benefits [1]. Despite concretes susceptibility to marine-based degradation phenomena [2], and much of the world’s marine infrastructure being located in cool with freezing climatic zones (annual average temperature < 10°C and average summer temperature generally < 20 °C) [3], research on the development of bacteria-based self-healing concrete has been largely restricted to room temperature freshwater studies [4-14]. The objective of the current project was, therefore, to develop a cost-effective bacteria-based self-healing cementitious composite for application in low-temperature marine environments. The current thesis charts the development of this composite. In Chapter 2 the autogenous healing capacities of ordinary Portland cement (OPC) and blast-furnace slag (BFS) cement mortar specimens submerged in fresh and seawater, are visually quantified and characterised. The BFS cement specimens healed all crack widths up to 104 µm, and OPC specimens healed all crack widths up to 592 µm, after 56 days in seawater. BFS cement specimens healed all crack widths up to 408 µm, and OPC specimens healed all crack widths up to 168 µm, after 56 days in freshwater. OPC specimens in seawater displaying the higher crack healing capacity also demonstrated considerable losses in compressive strength. Differences in performance are attributable to the amount of calcium hydroxide in these mortars and specific ions present in seawater. Chapter 3 reports on the crack healing capacity of seawater submerged mortar specimens with the aid of a crack permeability test. Cracks of defined widths were created in BFS cement specimens allowing reference crack permeability values to be generated for unhealed-specimens against which healed-specimens were quantified. Specimens with 0.2 mm wide cracks demonstrated no water flow after 28 days submersion. Specimens with 0.4 mm cracks demonstrated decreases in water flow of 66% after 28 days submersion and 50 to 53% after 56 days submersion. Chapter 4 presents a modified permeability test for generating crack permeability data for cementitious materials. To gauge for any improvement both the modified and unmodified tests were tested and compared. Cracks were generated in mortar specimens using both tests, the accuracy of these cracks was analysed through stereomicroscopy and computer tomography (CT), and the water flow through the cracks determined. Reduction factors and crack flow models were generated, and the accuracy and reliability of the predictions assessed. All of the models had high predictive accuracies (r2 = 0.97-0.98), while the reliability of these predictions was higher for the models generated with the crack width analysis through stereomicroscopy. The cracks generated by the modified test were more accurate (within 20 µm of the desired crack widths) than those of the unmodified test. The modified test was 30% quicker (10 hours for twenty-one specimens) than the unmodified test at generating the crack permeability data. Further, crack width analysis through stereomicroscopy is currently/generally quicker than analysis through CT. Chapter 5 presents a bacterial isolate and organic mineral precursor compound, as part of a cost-effective healing agent for low-temperature marine concrete applications. Organic compounds were screened based on their cost and concrete compatibility, and bacterial isolates based on their ability to metabolise concrete compatible organic compound and to function in a low-temperature marine concrete crack. Magnesium acetate was the cheapest organic compound screened, and when incorporated (1% of cement weight) in mortar specimens had one of the lowest impacts on compressive strength. Bacterial isolate designated psychrophile (PSY) 5 demonstrated very good growth under saline (3%), high pH (9.2), low-temperature (8ºC) conditions, with sodium lactate as an organic carbon source; and good growth at room temperature using magnesium acetate as an organic carbon source. Further, PSY 5 also demonstrated good spore production when grown on monosodium glutamate at room temperature. Chapter 6 presents a bacteria-based bead for realising self-healing concrete in low-temperature marine environments. The bead, consisting of calcium alginate encapsulating bacterial spores and mineral precursor compounds, was assessed for: oxygen consumption, swelling, and its ability to form an organic-inorganic composite in a simulative marine concrete crack solution (SMCCS) at 8ºC. After six days in the SMCCS, the bacteria-based beads formed a calcite crust on their surface and calcite inclusions in their network, resulting in a calcite-alginate organic-inorganic composite. The beads swell by 300% to a maximum diameter of 3 mm, while theoretical calculations estimate that 0.1 g of the beads are able to produce ~1 mm3 of calcite after 14 days submersion. Swelling and the formation of bacteria induced mineral precipitation providing the bead with considerable crack healing potential. It is estimated, based on the bacteria-based beads costing roughly 0.7 €.kg-1, that bacteria-based self-healing concrete made using these beads would cost 135 €.m-3. Chapter 7 presents a bacteria-based self-healing cementitious composite for application in low-temperature marine environments. The composite was tested for its crack healing capacity with the water permeability test presented in Chapter 4, and for its strength development through compression testing. The composite displayed an excellent crack healing capacity, reducing the permeability of cracks 0.4 mm wide by 95%, and cracks 0.6 mm wide by 93%, following 56 days submersion in artificial seawater at 8ºC. Some conclusions were drawn based on the results obtained during the development of the bacteria-based self-healing cementitious composite: • Visual crack closure is not a measurement for the regain of functional properties such as strength. Visual crack closure, therefore, should only be conducted as a complementary method when measuring the regain of such a property. • The capacity of a cementitious material to heal a crack depends on the width of the crack, thermodynamic considerations, the presence of water and the amount of ions available in the crack. Autogenous crack healing for seawater submerged cementitious materials is principally attributable to the precipitation of aragonite and brucite in the cracks. • The crack healing capacity of a bacteria-based cementitious composite is directly related to the amount of organic carbon available to the bacteria, and so the cheaper the organic mineral precursor compound, the cheaper the bacteria-based self-healing technology in general. Further, the compound must not have an adverse effect on concrete properties when included and must be readily metabolised by the bacteria as part of the healing agent. Magnesium acetate, in the current study, best balanced these criteria making it a good candidate as the organic mineral precursor compound for the healing agent. • A large number of specimen replicates (≥ 7) are required to generate reliable crack permeability data, and hence to quantify the crack healing capacity of cementitious materials through their functional water tightness. • The bacteria-based self-healing cementitious composite displayed an excellent crack healing capacity, reducing the permeability of cracks 0.4 mm wide by 95% and cracks 0.6 mm wide by 93%, following 56 days submersion in artificial seawater at 8ºC. This crack healing capacity was attributable to: mineral precipitation as a result of chemical interactions between the cement paste and seawater; bead swelling; magnesium-based precipitates as a result of chemical interactions between the magnesium of the beads and hydroxide ions of the cement paste; and bacteria-induced mineral precipitation. • The 28-day compressive strength of mortar specimens incorporated with beads was 55% of plain mortar specimens. Reducing the amount of bacteria-based beads will likely increase the compressive strength of the bacteria-based self-healing cementitious composite. Such a reduction, given the swellability of the beads, may have relatively little impact on the healing capacity of the composite. • The bacteria-based self-healing cementitious composite shows great potential for realising self-healing concrete in low-temperature marine environments, while the organic-inorganic healing material formed by the composite represents an exciting avenue for self-healing concrete research. I hope that the work presented herein provide a valuable reference for those interested in bacteria-based self-healing concrete, particularly for application in marine environments, and more generally for those interested in the wider field of self-healing materials research.