Abstract: Brackish water represents a potential supplementary water source for agriculture in the arid regions of Northwest China, but its salinity restricts direct utilization. The development of integrated desalination equipment requires evaluation of salt removal, energy consumption, and operational stability beyond small laboratory systems. This study developed brackish water desalination equipment based on membrane capacitive deionization (MCDI) with a bipolar electrode configuration and investigated its performance under different operating conditions.A laboratory desalination unit was constructed by stacking activated carbon electrodes with anion- and cation-exchange membranes. The effects of average voltage per electrode pair (1.0, 1.2, 1.5, and 1.8 V), circulating flow rate (1, 2, and 3 L/min), and initial salinity (0.5, 1, 2, and 4 g/L) were evaluated using synthetic NaCl solutions. The design was subsequently scaled up to 300 electrode pairs and integrated with pretreatment, water delivery, and monitoring and control units. A programmable logic controller recorded flow rate and conductivity during operation. Equipment performance was evaluated at influent flow rates of 250, 500, and 1000 L/h using the same voltage levels and initial salinities. Each equipment operating condition was tested in triplicate. Salt adsorption capacity (SAC), expressed as the mass of salt removed per unit electrode mass, was used to quantify desalination performance. Energy consumption and regeneration by polarity reversal were also assessed.Voltage was the operating parameter most strongly associated with SAC in the laboratory tests (Pearson's r = 0.749), whereas the correlation with flow rate was weak and negative (r = −0.094). These trends were also observed in the equipment tests. Increasing the average voltage per electrode pair from 1.0 to 1.8 V increased SAC by 14.44%–40.85% across the tested salinities. Increasing the influent flow rate from 250 to 1000 L/h reduced SAC after 10 min of adsorption by 15.58%–31.53%. At 250 L/h, the corresponding SAC values after 10 min were 17.06, 24.71, 30.95, and 37.31 mg/g for initial salinities of 0.5, 1, 2, and 4 g/L, respectively. Thus, lower flow rates and higher electrode voltages favored salt adsorption within the investigated operating range.Energy analysis demonstrated the need to balance salt removal against electricity consumption. At 250 L/h, increasing the average electrode voltage from 1.0 to 1.8 V increased unit energy consumption by 83.72%–132.68%, exceeding the relative gains in SAC. At a fixed electrode voltage, increasing initial salinity from 0.5 to 4 g/L reduced unit energy consumption by 27.45%–44.87%. For the 4 g/L solution, increasing the electrode voltage from 1.5 to 1.8 V improved SAC by only 0.79% but increased unit energy consumption by 45.85%. Separate measurements of volumetric energy consumption at a total equipment operating voltage of 350 V and a flow rate of 250 L/h yielded 0.881 and 0.980 kWh/m³ for initial salinities of 2 and 4 g/L, respectively. Increasing the total operating voltage to 400 V raised these values to 2.053 and 3.654 kWh/m³, respectively.Cycling tests were conducted using a 0.5 g/L NaCl solution at an average electrode voltage of 1.8 V and a flow rate of 250 L/h. Each cycle comprised 10 min of adsorption followed by desorption through polarity reversal. Over 25 adsorption–desorption cycles, conductivity recovery after desorption remained at least 98.41%, indicating effective regeneration and stable cycling behavior over the test period. The integrated equipment retained the principal voltage and flow responses observed in the laboratory unit. These findings provide an experimental basis for scaling up MCDI systems with bipolar electrode stacks and selecting operating conditions that balance salt adsorption, treatment flow rate, and energy consumption for brackish water utilization in arid agricultural regions.