Thermogalvanic cells convert waste heat directly to electric work. There is an abundance of waste heat in the world and the opportunity represented by thermogalvanic cells may be underused. We discuss theoretical tools that can help us understand and therefore improve the cell performance. One theory is able to describe all aspects of the conversion, nonequilibrium thermodynamics. We recommend to use the theory with operationally defined, independent variables, as others have done before. These describe well-defined experiments. Three invariance criteria can be used as a basis: of local electroneutrality, of entropy production invariance, and of emf's independence of the frame of reference. Alternative formalisms are using different sets of variables, ionic or neutral components. We show that the heat flux is not the same in the two formalisms and derive a new relationship between the heat fluxes. The heat flux enters the definition of the Peltier coefficient and is essential for the understanding of the Peltier heat at the electrode interfaces and of the Seebeck coefficient of the cell. The Soret effect can occur independent of any Seebeck-effect, but the Seebeck effect will be affected by the presence of a Soret effect. Common misunderstandings are pointed out. Peltier coefficients are needed for interpretation and design of measurements.