Page 22

EETE SEPTEMBER 2012

DESIGN & PRODUCTS ENERGY HARVESTING where rteG,thermal is the thermal generator module was held in compression be- resistance of the thermoelectric elements, tween the heat sink and the strap base with two Hsr is the thermal resistance from the hot stainless steel bolts fitted with insulating phenolic source to the hot side of the thermoelectric washers. thermal pads on both sides of the teG elements and Csr is the thermal resistance were used to improve thermal resistance at the from the cold source to the cold side of the hot (strap) and cold (heat sink) interfaces. the thermoelectric elements. clamping strap base was divided into three identi- Figure 2 represents the impact on cal sections that form a compression fit around performance that different thermal load the pipe when bolted together. a series of lab ratios have on the power output. For most tests were conducted with this design that mim- Fig. 1: A 10-inch diameter thermoelectric applications, designing icked varying operating conditions throughout the for a thermal load resistance ratio of one implementation of the EverGen year. the test assembly was made from a section ensures the best performance possible. In PowerStrap energy harvester from of 10” diameter steel pipe that was capped at the case of the everGen Powerstrap, the Marlow. one end and filled with oil. Submersible heaters, best performing natural convection heat sink was chosen, based on orientation, size, cost and manufacturing constraints. Computational fluid dynamics (CFd) software was used to aid in the heat sink design optimization. next, the teG devices were designed using Marlow’s proprietary teG software to match the thermal resistivity of the natural convection Fig. 2: Impact of thermal load Fig. 3: Impact of electrical heat sinks under pure natural convection conditions. matching on the EverGen PowerStrap load matching on the EverGen the electrical system optimization is analogous to design. PowerStrap. the thermal system. For maximum power transfer, the internal electrical resistance of the power source must match attached to an electronic temperature controller, were used to the electrical resistance of the load being powered. In this case, control test assembly wall temperature. the Powerstrap was the electrical load ratio (n) is defined as clamped to the exterior of the test assembly, with non-setting where rload is the electrical resistance of the load being pow- thermal mastic applied between the pipe wall and the strap ered and rteG,electrical is the electrical resistance of the teG base to aid in heat transfer. during testing, ambient temperature module under operating conditions. around the test assembly was altered to reflect seasonal chang- Figure 3 highlights why this is a particularly important con- es. Both natural convection and forced convection up to 6.5 sideration when designing thermoelectric power generation mph were studied. omega oM-420 data acquisition equipment systems. In reality, both electrical and thermal characteristics of was used to collect temperature, voltage and current measure- the teG are interrelated with the thermal resistance of the teG ments during testing. Figure 4 is an expanded view sketch that being affected by the electrical load connected to the TEG. In depicts thermocouple placement on the test assembly. read- real world applications, where operating conditions and loads ings were collected and recorded in two second intervals. vary, it would be very difficult to always ensure proper load matching across all operating points due Test results to temperature dependant properties the results of this testing, compared against model of the teG. Fortunately, commercially predictions, are shown in Figure 5 for two different available maximum power point tracking pipe temperatures covering a wide range in ambi- (MPPt) controllers originally designed for ent conditions. From the data, it is obvious that the the solar industry can also perform this everGen Powerstrap performance is maximized function for thermoelectric systems. In when ambient temperatures are the coldest. this cases where hybrid solar/thermoelectric is to be expected since thermoelectric efficiency is systems are employed, a single MPPt greater for larger temperature differentials. In real controller accommodates both. the only Fig. 4: Expanded view sketch of world operation, this means that the Powerstrap design requirements are that the teG the thermocouple placement on performance will be maximized during the colder system voltage and current outputs for months of the year. such performance makes this the test assembly. the operating range meet the input re- product a natural complement to solar cells, which quirements of the MPPt controller. usually perform poorly during the winter months. Another key point is that there is significant perfor- Test setup mance increase, by as much as 40%, when typi- Figure 1 shows a photograph of a 10- cal outdoor wind conditions are accounted for. For inch diameter everGen Powerstrap. the applications requiring higher power levels, multiple unit was designed for outdoor use, 120°C units can be employed. the data also shows that the operation in a vertical orientation for an model predictions close well with experimental data. industrial exhaust pipe. twelve identical By expanding the model to include different pipe teG and heat sink assembly sections Fig. 5: The EverGen PowerStrap temperatures and diameters, performance under dif- were spaced evenly around the perimeter ferent operating scenarios can be predicted. test results. of the strap base. each thermoelectric 22 Electronic Engineering Times Europe September 2012 www.electronics-eetimes.com


EETE SEPTEMBER 2012
To see the actual publication please follow the link above