We examine the convergence of third-order structure function expressions derived to measure the rate of turbulent energy cascade within the solar wind using Advanced Composition Explorer observations from 1 AU over the years 1998 through 2007. We find that a minimum of a year of data is normally required to get good convergence and statistically significant results. We then apply these findings to 10 years of observations spanning both solar minimum and solar maximum conditions. We compare the computed energy cascade rates with previously determined rates of proton heating at 1 AU as determined from the radial gradient of the proton temperature to be proportional to the product of wind speed and proton temperature. We find good agreement with a moderate excess of energy within the cascade that is consistent with previous estimates for thermal electron heating in the solar wind. In keeping with earlier analyses of the dissipation spectrum, we postulate that electron heating by the turbulent cascade is less than and at most equal to the rate of proton heating.