The lack of energy closure has been a longstanding issue with Eddy Covariance (EC). Multiple mechanisms have been proposed to explain the discrepancies in energy balance including diurnal energy storage changes, advection of energy, and larger scale turbulent processes that cannot be resolved by field EC. To investigate the energy balance issue, we used a year of data from paired EC towers in irrigated sugarcane in Maui, Hawai’i, USA. The towers were in identical crops and cultivation practices and had similar climate with the notable exception of wind. One tower was in a location where nearby orographic features funneled Trade Winds, resulting in sustained, continuous turbulence. The other was in a leeward location with less turbulence, particularly at night (u*). We found significantly improved closure (8.5–10%) at both sites using daily sums of Available Energy in closure regressions as opposed to 30 min data, illustrating the importance of storage terms. The energy budget closed for both fields when only days with continuous turbulence (all 30 min u* > critical u*) were considered, with significantly larger uncertainty in the leeward field (±13%) due to the small number of days (n = 13) with this condition. Significant energy imbalance appeared in both fields with even 30 min of subcritical turbulence in a day, and each field had different turbulence-closure patterns. Closure with continuous turbulence was sensitive to choice of critical u*; an arbitrary u* of 0.1 m s−1 resulted in non-closure. The results show the value of paired EC towers in contrasting turbulence conditions to assess energy budget closure.