Large amounts of low-grade heat (T < 100 °C) are readily available from a wide range of sources, such as industrial facilities, geothermal wells, and solar thermal systems. The potential benefits of harvesting low-grade heat sources to produce power include reduced carbon footprints and increased system efficiencies of thermal power generation systems. Thermally regenerative ammonia batteries (TRABs) are one of the more promising technologies in the space, but like many other emerging technologies, they still lack the power density needed to be economically viable. Though bimetallic TRAB (B-TRAB) systems have increased TRAB power densities by using two dissimilar metals to dramatically increase cell potentials, they suffer from poor coulombic efficiencies due to their reliance on metallic copper in the presence of dissolved ammonia. A typical bimetallic TRAB uses four steps for its cycle operation: (1) high-voltage electrochemical discharge, (2) thermal ammonia separation, (3) low-voltage electrochemical charge, and (4) thermal ammonia separation. To address B-TRAB efficiency issues, we reduced the number of steps needed to complete a battery cycle to three by removing a thermal charging step and modified the electrolyte composition to favor the Cu(I, II) reaction at the positive electrode. Cycle tests with the new 3-step B-TRAB increased coulombic efficiency to 85%, nearly double the coulombic efficiency of the 4-step B-TRAB. Likewise, the new system had an average discharge power density of 15.55 mW cm^-2, 25% larger than the 4-step B-TRAB, with a net energy density of 0.92 Wh L^-1 at a current density of 10 mA cm^-2. These modifications to the B-TRAB concept collectively increased multiple performance metrics, further increasing the favorability of TRABs as competitive low-grade waste heat harvesting devices.