Solar thermal energy conversion to electrical power

Applied Thermal Engineering 70 (2014) 675e686

Contents lists available at ScienceDirect

Applied Thermal Engineering journal homepage: www.elsevier.com/locate/apthermeng

Solar thermal energy conversion to electrical power lez b, Luc Fournier c, Re mi Pelletier c, Juan C. Sandoval V. b, Anh-Khoi Trinh a, Ivan Gonza c , d, * de ric J. Lesage Fre a

McGill University, 845 Sherbrooke St W, Montreal H3A 0G4, Canada n de Ingeniería y Tecnología, Universidad Tecnolo gica de Cancún, Carretera Cancún-Aeropuerto, Km. 11.5, S.M. 299, Mz. 5, Lt 1, Cancún 77500, Divisio Quintana Roo, Mexico c Cegep de l'Outaouais, 333 boul. de la Cit e-des-Jeunes, Gatineau J8Y 6M4, Canada d Departement d'informatique et d'ing enierie, Universit e du Qu ebec en Outaouais, 101 rue Saint-Jean-Bosco, Gatineau J8Y 3G5, Canada b

h i g h l i g h t s Solar radiation maintains a thermal tension which drives an electromotive force. Voltage, current and electric power are reported and discussed. Theoretical optimal thermoelectric conversion predictions are presented. Theory is validated with experimentally measured data.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 January 2014 Accepted 27 May 2014 Available online 9 June 2014

The conversion of solar energy to electricity currently relies primarily on the photovoltaic effect in which photon bombardment of photovoltaic cells drives an electromotive force within the material. Alternatively, recent studies have investigated the potential of converting solar radiation to electricity by way of the Seebeck effect in which charge carrier mobility is generated by an asymmetric thermal differential. The present study builds upon these latest advancements in the state-of-the-art of thermoelectric system management by combining solar evacuated tube technology with commercially available Bismuth Telluride semiconductor modules. The target heat source is solar radiation and the target heat sink is thermal convection into the ambient air relying on wind aided forced convection. These sources of energy are reproduced in a laboratory controlled environment in order to maintain a thermal dipole across a thermoelectric module. The apparatus is then tested in a natural environment. The novelty of the present work lies in a net thermoelectric power gain for ambient environment applications and an experimental validation of theoretical electrical characteristics relative to a varying electrical load. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Solar thermoelectric generator Solar evacuated tube Thermoelectric module Electrical load

1. Introduction Innovations in the field of solar energy conversion to electricity using the thermoelectric effect have increased in recent years. The main body of these works focus on hybrid photovoltaicethermoelectric systems [e.g., Refs. [1e7]] and on devices solely relying on the thermoelectric effect e commonly referred to as Solar Thermoelectric Generators (STEGs) [e.g., Refs. [8e13]].

partement d'informatique et d'inge nierie, Universite * Corresponding author. De bec en Outaouais, 101 rue Saint-Jean-Bosco, Gatineau J8Y 3G5, Canada. du Que Tel.: þ1 819 595 3900; fax: þ1 819 773 1638. E-mail addresses: [email protected], [email protected] (F.J. Lesage). http://dx.doi.org/10.1016/j.applthermaleng.2014.05.088 1359-4311/© 2014 Elsevier Ltd. All rights reserved.

The goal of the thermoelectricephotovoltaic hybrid investigations is to convert excess unwanted heat resulting from the thermophotovoltaic effect (energy not absorbed by the photovoltaic cell's band gap is converted to heat [e.g., Ref. [14]]) into electricity. This is accomplished using thermoelectric modules which have embedded doped semiconductors capable of generating an electromotive force from a thermal differential. This thermoelectric phenomenon and its application to other waste-heat recovery uses are described in detail in Refs. [15e20] among others. The present work focuses on the thermoelectric conversion of solar energy without the use of photovoltaic materials. In this case, an asymmetric thermal field ideally creates a thermal dipole across a devices' embedded thermoelectric module in which solar radiation provides the heat source. Early investigations in the field of Solar Thermoelectric Generators (STEGs) used a radiation concentration

676

A.-K. Trinh et al. / Applied Thermal Engineering 70 (2014) 675e686

Nomenclature A I I* L m k P P* q q_ R R* T TEM V

cross-sectional area, m2 electrical current, A I/Iopt thermopellet length, m mass, kg thermal conductivity, W/m K power, W P/Popt heat transfer rate, W thermal energy generation per unit volume, W/m3 electrical resistance, U R/Ri, temperature, K thermoelectric module voltage, A

technique via heated aluminium blocks. For example, Goldsmid et al. [21] injected heat to the hot side of a thermoelectric module by thermal conduction through an aluminium block exposed to the sun. The thermal differential across the module was maintained by dissipating heat through natural convection into the ambient air. In their study, they found that a greater concentration of solar energy and a greater conversion efficiency was needed to improve system thermal input. To this end, Rowe [22] investigated the optical efficiency of a siliconegermanium thermoelectric module, Chen [23] developed a thermodynamic model to investigate the optimal performance of STEGs, Lenoir et al. [24] evaluated the electrical properties of STEGs based on the mineral skutterudite for possible aerospace applications, Bomberger et al. [25] investigated the effects of varying thermal input conditions on thermoelectric module performance for solar and other applications, Jang and Tsai [26] analysed the optimal module spacing for solar applications, and Weinstein et al. [27] showed that thin-film STEGs have a similar conversion efficiency to that of existing bulk material STEGs. Furthermore, design optimization of a STEG's embedded thermoelectric modules has been investigated by Wang et al. [28], Inagoya et al. [29], Bhardwaj [30], Xiao et al. [31], Al-Merbati et al. [32] and Ali et al. [33] among others. In an effort to enhance the thermal input due to the harnessed solar radiation, Vatcharasathien et al. [34] used compound parabolic collectors to increase the heat source to a series of sixteen thermoelectric modules and a refrigeration unit as a heat sink. The energy consumption however of the refrigeration unit offset any thermoelectric power gain. Similarly, Mgbemene et al. [35] used compound parabolic concentrators to enhance the solar radiation heat source to the thermoelectric module but used a simple fan and ambient air for cooling purposes. The thermoelectric conversion efficiency of their device was reported to attain a maximum of 0.24%. Solar concentrators can also be used to supply heat to a thermoelectric cogeneration system (TCS) [e.g., Refs. [36,37]]. Yazawa et al. [38,39] used Fresnel lenses to further concentrate the solar radiation heat applied to the module and pumped cold water to the cold side of the module in order to enhance thermal transport of the heat sink. In doing so they effectively increased the thermal dipole yet the cold water flow refrigeration cost offset the thermoelectric power gain. Furthermore, the disadvantage to concentrator lenses is that they increase the total area of operation of the device thereby compromising the system's output per unit area. Since thermoelectric power output increases exponentially with respect to the thermal dipole of the asymmetric thermal field in

V*

V/Vopt

Greek letters a seebeck coefficient, V/K ap,n ap an DT TH TC, K r electrical resistivity, U m Subscripts C cold junction H hot junction i internal L load oc open circuit opt optimal n N-type material p P-type material

which it is subjected to [e.g., Ref. [40]], He et al. [41,42] stocked solar radiation in evacuated solar tubes in order to further increase the thermoelectric module's heat source side temperature. Solar tubes effectively capture solar radiation by heating a gas embedded in the tube which is insulated by a vacuum double walled outer encasing. A performance evaluation of these tubes is provided in Refs. [43e45] showing that they are capable of capturing solar radiation up to an 80% efficiency. With these tubes, He et al. [41,42] channelled solar thermal energy from the natural Sun to the hot side of a single thermoelectric module. The thermal differential across the modules was maintained by pumping cold water to the heat sink side of the module. In their study, a solar energy to electrical energy conversion efficiency ranging from 0.6% to 1.5% was reported without accounting for the pumping penalty associated with the work done to channel cold water to the heat sink. Similarly, Zhang et al. [46] used solar tubes to apply heat to a set of thermoelectric modules refrigerated with cold water forced convection providing a hot water by-product. In their configuration, the device's thermoelectric production partially offsets the pumping cost. The difficulties in the current state-of-the-art in solar thermoelectric generators lie in the power cost of the heat sink. The most promising heat source for STEG power generation reported in the afore mentioned literature is that which captures solar radiation with evacuated solar tubes. However, currently tested devices use water cooling systems at the heat sink which require costly hydraulic power. In an effort to alleviate this adverse pumping penalty, the present work presents an apparatus that effectively converts solar radiation to electricity without relying on an external water pump for cooling. In this investigation, solar energy is captured and stocked in a double walled vacuum insulated solar tube. This harnessed thermal energy is driven towards one side of a thermoelectric module by way of an embedded copper cylindrical tube. The cold side of the module is cooled by dissipating heat into the ambient air with the use of a CPU heat exchanger and a ventilation simulating wind assisted forced convection. The resultant thermal dipole generates an electromotive force in the embedded Bismuth Telluride Bi2Te3 semiconductors. This material is used in the present study since it has been shown to be the most effective thermal to electric conversion material for the present work's target temperature range of 40e120 C [e.g., Ref. [47]]. The apparatus is then used in a field test study in which the heat source and the heat sink are naturally occurring solar radiation and wind aided air convection respectively.


The present reports and discusses the potential to harness and convert solar and wind energy into electrical power for a net electrical gain when considering the device's entire thermal system. By including a detailed description of the measured heat source and heat sink conditions and the resultant thermoelectric characteristics, the present work presents novel STEG results when subject to a variable electrical load. This is relevant to STEG applications since power point tracking systems adjust the electrical circuit's load in order to maintain peak power operation under variable thermal input conditions [e.g., Refs. [18,40]]. For instance, maximum power point tracking circuitry is designed and applied to thermoelectric devices in Kim and Lai [48], Kim et al. [49], and Zhang and Chau [50]. A detailed description of the theoretical optimal thermoelectric characteristics relative to a varying electrical load is presented and validated. Furthermore, preliminary results of a field test in the naturally occurring solar and wind sources of the state of Quintana Roo Mexico are per gica de Cancun's campus formed on the Universidad Tecnolo grounds. This work is relevant to the scientific community investigating STEG applications to the physical ambient environment since it presents a significant improvement in STEG efficiency operating on simulated and naturally occurring solar and wind energy. This implies that there is no adverse pumping penalty resulting from hydraulic pumps to the heat sink. The STEG power output result of 4.82 W/m2 is reported and discussed in detail. 2. Thermoelectric generation

Electrical power generation by way of thermoelectric modules is made possible by the principle known as the Seebeck effect. Electricity is characterized as the flow of electrons in a circuit and occurs more easily in elements which hold only one valence electron in their outer orbits. This is due to the fact that valence electrons are loosely bound to the atom's nucleus and can therefore easily move to a neighbouring atom. The Seebeck effect is the phenomenon of a thermal dipole exciting such charge carrier mobility. Upon closing the electrical circuit, this ability to drive an electromotive force from a thermal differential results in a thermoelectric power production. However, elements that have only one valence electron, such as copper, are commonly also thermal conductors thereby nullifying the thermoelectric effect. This is due to the fact that a tendency towards temperature equilibrium amongst the materiel's extremes decreases the thermal dipole thereby decreasing charge carrier mobility. In order to minimise the thermal conductivity while maximising charge carrier concentration, elements are combined into crystalline structures in which atoms share their valence electrons. When a crystalline has multiple free electrons roaming from one atom to another, it is called an N-Type material since it is negatively charged. Similarly, when a crystalline is doped to have multiple electron “holes” e atoms charged positively due to an absence of electrons e it is called a P-Type material. When placing a conductor between two different temperature areas, the Seebeck effect is effectively a thermal energy transfer flow from the hottest area to the coldest in which electrical charge carriers within the conductors tend to flow in the same direction as the heat transfer, thus creating a flow of charge. The Seebeck coefficient is used to quantify this phenomenon. It represents the ratio of the electrical potential to the thermal differential, defined as [e.g., Ref. [51]]

a¼

V ; DT

and it is an intrinsic part of a material's electronic structure and its thermoelectric characteristics. It is important to note that for a single N-type material, the heat source drives negative charge carriers to the heat sink creating an overload of electrons without positive “holes” to move into. In order to circulate the charge carriers into an electric circuit, thermoelectric modules are designed to alternate positive and negative thermoelectric materials into thermocouples (thermally in parallel and electrically in series) in order to generate an electric circuit. The idealized case of an electric circuit resulting from a thermal dipole is illustrated in Fig. 1. 2.2. Optimal thermoelectric characteristics The thermoelectric phenomenon for target applications is commonly optimised through thermal system management of the available heat source and heat sink [e.g., Ref. [52]]. For example, Amaral et al. [20] reported thermoelectric power enhancement by way of turbulence regime management in inner channel flow for a liquid-to-liquid thermoelectric generator. Furthermore, thermoelectric power output is maximised through power point tracking means in which an optimal electrical load is identified for target working conditions [e.g., Ref. [18]]. The optimal thermoelectric characteristics for a thermoelectric module can be deduced from the thermoelectric power output calculation [e.g., Ref. [53]] resulting from the one-dimensional heat equation, such that

d dT k q_ ¼ 0 dx dx

2.1. Thermoelectric effect

(1)

677

(2)

for which k is the thermal conductivity of the semiconductor material. In order to discuss the theoretical thermoelectric characteristics of a conversion of thermal energy to electrical energy by way of the Seebeck effect, the heat equation is applied to an idealized case in which a thermal differential is acting on a pair of N-type and P-type materials which are electrically connected as illustrated in Fig. 2. The term q_ in Eq. (2) therefore represents the thermal energy converted to electricity by the Seebeck effect per unit volume. Applying Ohm's law, it is quantified as,

I2 q_ ¼

rL A

A$L

(3)

in which r is the material's electrical resistivity, A is the crosssectional area relative to current flow in the material, and L is the length of the material.

Fig. 1. Electric circuit resulting from an asymmetric thermal field idealized as a thermal dipole.

678


The thermoelectric power is therefore the difference in the heat flux at the heat source and the heat flux at the heat sink resulting in,

P¼

I 2 rL þ Iap;n DT: A

(8)

The term rL=A represents the electrical resistance of the thermoelectric material reducing Eq. (8) to,

I2 RL ¼ I2 Ri þ Iap;n DT

(9)

in which RL and Ri are the external electrical load resistance on the circuit and the internal electrical resistance of the material respectively. From Eq. (9), the closed circuit electrical current is,

I¼

ap;n DT RL þ Ri

(10)

and the electric potential across the electrical load is,

VL ¼ Fig. 2. Schematic representation of an electrically connected pair of N-type and P-type materials subject to a thermal differential.

RL ap;n DT : RL þ Ri

(11)

Combining Eqs. (8) and (10) yields the thermoelectric power production from a thermal differential,

P ¼ DT 2 a2p;n

RL ðRi þ RL Þ2

:

(12)

By applying boundary conditions in which TH and TC are the hot and cold junction temperatures respectively of the thermal dipole, applying Ohm's law, and assuming that the Thomson effect [e.g., Ref. [51]] is negligible, the one-dimensional heat equation generates the following temperature profile for a thermoelectric material of length L,

A derivation of thermoelectric power with respect to electrical load resistance shows that the theoretical peak power occurs upon electrical impedance matching. Therefore, the optimal electrical load which maximises power is equal to the internal electrical resistance reducing Eq. (12) to,

2 I2 r 2 I rL TH TC x þ TC : T ¼ x þ þ L 2kA2 2kA2

Popt ¼

(4)

In Eq. (4), A includes the cross-sectional areas of both the Ntype and P-Type materials. The principle of conservation of energy requires that at each junction the heat flux “in” be equal to the heat flux “out” plus the energy converted to electricity by the Seebeck effect. Since the Seebeck coefficient represents the ratio of the electric potential to the thermal differential as expressed in Eq. (1), the heat balance at the hot junction may be expressed as,

qH ¼ qH;out þ Iap TH Ian TH

(5)

in which an and ap are the Seebeck coefficients for the N-type and P-type materials respectively. It is also important to note that the sign for the terms containing the electric current in Eq. (5) are with respect to the current flow direction illustrated in Fig. 2. Applying this principle to Fourier's law of conduction to deduce qH,out from the temperature profile develops Eq. (5) into,

qH ¼

kA I 2 rL DT þ Iap;n TH L 2A

kA I 2 rL DT þ þ Iap;n TC : L 2A

4Ri

(7)

(13)

in which Popt is the optimal thermoelectric power corresponding to a peak power output. Similarly, load matching reveals the optimal electrical current to be,

Iopt ¼

ap;n DT 2Ri

(14)

and the optimal electric potential across the load as,

VL;opt ¼

ap;n DT : 2

(15)

In an effort to predict optimal thermoelectric characteristics for any thermoelectric material subject to a given thermal differential, the power output expressed in Eq. (12) is normalized by the peak power output of Eq. (13) such that P* represents the normalized power,

(6)

in which ap;n is the difference in the P-type and N-type material Seebeck coefficients and DT ¼ TH TC. Similarly, the heat transfer rate at the cold junction is found to be,

qC ¼

DT 2 a2p;n

P* ¼

4R*L 1 þ R*L

2

(16)

in which R*L represents the electrical load normalized by the material's internal electrical resistance. Eq. (16) is therefore the theoretical normalized thermoelectric power output of a material. Similarly, the optimal electrical current and the optimal electric potential across the load resistance


679

Fig. 3. Thermal differential across a thermoelectric module maintained by a solar tube and a heat sink.

(17)

electrical load is deduced from a measurement of the open circuit voltage Voc and knowledge of the electrical load upon closing the circuit. More precisely, the voltage across the thermoelectric module is measured upon opening the circuit. The circuit is then closed and the sum of the closed circuit voltages across the module and the electrical load is equated to the open circuit voltage. This yields a relationship between the internal electrical resistance and the measurable electrical characteristics,

(18)

Voc ¼ IRi þ VL :

normalize the electric current and the load voltage respectively reducing Eqs. (10) and (11) to,

I* ¼

2 1 þ R*L

and

VL* ¼

2R*L : 1 þ R*L

Furthermore, combing Eqs. (17) with (18) yields the normalized current in terms of the normalized electric potential,

I * ¼ 2 VL* :

(19)

In addition to presenting a new experimental method regarding the conversion of solar radiation to electrical energy, the present works aims to validate the normalized terms expressed in Eqs. (16)e(19) using the recorded data. In order to validate these results with experimentally measured power evolution curves, it is necessary to measure the internal electrical resistance of the material. This is done in accordance to electrical circuit theory [e.g., Ref. [54]] in which the internal

(20)

With Eq. (20) the internal electrical resistance of the module is measured for each test cases. 3. Experimental setup 3.1. Concept The goal of the present study is to capture solar radiation in the form of a heated gas within a solar tube and to use this thermal energy as a heat input to one side of a thermoelectric module. In order to investigate the device's ability to maintain a thermal dipole for electric power generation without the use of a water refrigeration pump, the heat sink used is ambient air with ventilation

Fig. 4. Solar thermoelectric test apparatus.

680


simulating Wind energy. The concept of the present investigation is presented in Fig. 3 illustrating the target heat source and the target heat sink. 3.2. Test apparatus In order to verify the efficiency of solar energy conversion to electricity by way of the thermoelectric effect, a test stand illustrated in Fig. 4 is commissioned using a combination of 10 bulbs to replicate the Sun's irradiance. The resulting test lamp is referred to as the “Sunlamp” in the present document. More precisely, the Sunlamp consists of two sets of four 65 W Xenon crystal Eurolite 9005XB bulbs and two 400 W FloraSun MH bulbs combined and placed within a half-opened aluminium cylindrical casing. The Xenon bulbs are arranged into two vertical columns and are each subject to 11 V using a 36 V LabVolt power supply per bulb. The FloraSun bulb positioned in the upper section of the Sunlamp operates at 120 V while the FloraSun bulb positioned in the lower portion of the Sunlamp operates at 110 V. For each column of Xenon bulbs, each bulb is a vertical distance of 11 cm from each other and a horizontal distance of 10 cm from the nearest side of the cylindrical casing. The two FloraSun bulbs are placed in a central vertical column arrangement an equidistance between the Xenon bulbs. The Sunlamp casing measures 0.330 m in width and 0.595 m in height. Furthermore, the solar tube harnessing the Sunlamp's light energy is placed 12 cm in front of the Sunlamp and 16 cm above the ground whereas the Sunlamp is positioned 8 cm above the ground. A copper tube is inserted into the solar tube to assist in the thermal transport to a Bismuth Telluride (Bi2Te3) thermoelectric module which is positioned atop the copper tube. A CPU heat exchanger with an integrated ventilation sits atop the module acting as a heat sink and has 12 V applied to it. 3.3. Component specifications With this apparatus, radiation emitted from the Sunlamp is absorbed by a Clean Republic double walled solar tube measuring 0.485 m in length with a vacuum seal between the walls for thermal insulation. The inner diameter of the tube measures 43 mm and its outer diameter measures 58 mm. In the interest of further focussing solar radiation emitted by the lamp, an aluminium hemispherical reflector of 0.45 m in length is placed 35 cm behind

the solar tube with its concavity facing the solar tube. Furthermore, a copper cylinder is hollowed out to form a copper tube which is open on one end and closed at the other. This copper tube is 44.3 cm in length, has a 43.0 mm outer diameter, an outer wall thickness of 3.95 mm, and a base thickness of 13.0 mm. This copper tube is placed inside the solar tube effectively enhancing thermal transport the length of the solar tube to the thermoelectric module (TEM). Aluminium foil sheets of 0.8 mm in thickness are wrapped around the copper tube in order to increase thermal contact between adjacent materials. Heat harnessed at the lower end of the embedded copper tube is transferred to the graphite coated hot side of a TEG1-1263-43 thermoelectric module. The thermoelectric module is of dimensions 30 30 3.75 mm. A heat sink on the cold side of the TEM is provided by a Cooler Master Hyper 212 CPU Heat pipe exchanger which transports heat away from the upper end of the apparatus. The device’s fan measures 120 120 25 mm and provides up to 83 cubic feet per minute of ventilation. In an effort to estimate the air speed provided by the ventilator to the CPU cooler, an Omega HHF-SD1 Anemometer is used to map the air speed between the ventilator and the CPU cooler using 11 11 grid nodes. The mapping is provided in Fig. 5 in which the average air speed is calculated to be 1.72 m/s. The results illustrated in Fig. 5 show a non-uniform air speed projection onto the CPU cooler. The average air speed of 1.72 m/s is of the same order of magnitude as the 2.02 m/s average wind speed of the field test application presented in Section 4.9. For this reason, the artificial air speed is deemed suitable for the purpose of this investigation's laboratory controlled experimentation of the STEG. In order to improve thermal transport on both sides of the module, BD Hardware House Gold Thermal Grease (1.8 W/m-K thermal conductivity) is applied to each side of the module and a 1 kg mass is placed on top of the CPU diffuser to enhance contact between the different components and for device stability. In the field test application of the forthcoming Section 4.9, the CPU heater is fixed to the TEM using rigid Plexiglas and bolts as depicted in Fig. 6 Omegatherm 201 thermal compound (2.24 W/m-K thermal conductivity) is applied between the copper tube and the sheets of aluminium in order to increase thermal contact between the solar tube and the embedded copper tube. The principle components of the apparatus are presented in Fig. 7. 3.4. Electrical circuit

Fig. 5. Air speed generated by the fan at the contact surface with the CPU cooler.

-Potvin [18], the thermoelectric As detailed in Lesage and Page effect is optimised electrically by applying the optimal electrical load to the circuit for maximum thermoelectric power production. To this end, the present study uses Datastudio software package to relay measurements from the embedded sensors to the computer for data acquisitioning. More specifically, the voltage is measured at a fixed known electric resistance of 0.4 U in order to accurately measure the resultant electric current from the thermoelectric potential. The electric load resistance is varied using a Ten Turn Wirewound Potentiometer (Model 860) connected to a variable power motor controlled by a LabVolt e 4194 power supply resulting in an electrical load resistance rate of increase of 0.703 ± 0.0014 U/s. A second voltage reading is taken at the variable electrical load resistance in order to measure its rate of change which is made possible by the electric current measurement. A voltage reading is placed at the thermoelectric module in order to measure the thermoelectric power generated as a function of the variable electrical load. A schematic representation of the electrical circuit is presented in Fig. 8.


681

4. Results and discussion 4.1. Sunlamp spectrum irradiance In order to measure the wavelength spectrum of the Sunlamp, an ASD Inc. FieldSpec3 350-2500 spectroradiometer is used to measure the Sunlamp's luminance characteristics. The results illustrated in Fig. 9 show that the peak spectrum irradiance tendencies of the Sunlamp are of the same order of magnitude as that of the natural light spectrum irradiance tendencies of the Sun as reported by the American Society for Testing and Materials (ASTM) and published in Gueymard et al. [55]. More specifically, the spectrum irradiance of the Sunlamp is measured to increase from negligibly small to a peak irradiance of the order of 1 W/m2nm for wavelengths between 500 and 1000 nm, followed by a decrease to negligibly small irradiance in the upper wavelength range. In the context of the present study in which it is the solar tube's ability to capture solar radiation and to use it as a heat source for the device's embedded thermoelectric module, the fact that the solar tube is selective in its wavelength absorption and that the Sunlamp's wavelength spectrum is less homogeneous than that of the Sun's (the Sunlamp's wavelength spectrum contains more peaks and troughs), it is duly noted that the solar tube's light energy absorption is less efficient when operating with the Sunlamp than with the natural Sun. Even still, the Sunlamp provides controllable and repeatable experimental conditions making it ideal for investigating the thermoelectric conversion of solar energy to electricity. 4.2. Sunlamp irradiance Fig. 6. CPU Cooler fixation to the TEM for the purpose of a field test application.

As previously discussed, the target energy source to be converted to usable heat for thermoelectric power production is solar radiation. In an effort to further validate the Sunlamp's simulation of the Sun's radiation, the power intensity of the Sunlamp is compared to that of the Sun with the use of a 4 3 grid of sensors. This is accomplished using a Newport Oriel 91150V Reference Cell. This solar simulator calibration instrument consists of a

Fig. 7. Components of the solar thermoelectric test apparatus.

682


Fig. 8. Schematic representation of electric circuit including the variable load resistance.

2 2 cm monocrystaline silicon solar cell. The device provides measurements in Sun units in which a single unit is equal to 1000 W/m2. Each point of the grid is separated by 5 cm in the horizontal direction and by 12 cm in the vertical direction. The results illustrated in Fig. 10 show that the Sunlamp emits a greater irradiance in the upper section. The average of the irradiance measurements at the grid sensory points is found to be 804.1 W/m2. Measurements of the Sun's irradiance for local conditions (45 280 3800 N parallel of latitude and 50 m above sea level) yield values ranging from 812 W/m2 to 852 W/m2 within the period of July 22nd 2013eAugust 12th 2013. The Sunlamp therefore provides a solar irradiation of the same order as to that of the natural Sun for local conditions. It is deemed appropriate for use in the present study as a controlled solar energy source for the target application of thermal to electrical energy conversion.

reading is noted T1. The second thermal sensor is placed at a depth of 30 cm and its temperature reading is noted T2. In the test case using the natural Sun, the Sun's irradiance is considered as direct normal irradiance. For each test case : 1) capturing light energy in the solar tube with the natural Sun; and 2) capturing light energy in the solar tube with the Sunlamp, the temperature readings T1 and T2 are recorded over 45 min of light exposure. The results illustrated in Fig. 11 show that T1 and T2 readings obtained when operating with the Sunlamp are within 4% and 1% respectively of the corresponding values obtained at the same sensor locations when operating with the natural Sun. Furthermore, the rate of increase in temperature when operating with the Sunlamp at sensors T1 and T2 are within 7.3% and 2.8% respectively of the corresponding rate of temperature increase at the same sensor locations when operating with the natural Sun. In light of these results, the Sunlamp is deemed suitably calibrated to the Sun's electromagnetic radiance for the purpose of the present study's thermoelectric investigation.

4.3. Solar tube thermal energy absorption 4.4. Solar tube efficiency Thermal energy harnessed by the solar tube is measured by filling it with 600 ml of water and placing it 12 cm in front of the Sunlamp and then repeating the experiment by exposing it to the naturally occurring luminance of the Sun. In this way, temperature readings and knowledge of the specific heat of water make it possible to measure the solar tube's energy intake with respect to the Sunlamp and with respect to the natural Sun. To this end, two Omega HH804U thermal detectors are placed inside the tube. The first temperature sensor is placed at a depth of 15 cm from the upper opening of the tube and its temperature

The thermal network used to describe the thermal transport of the solar tube to the thermoelectric model is illustrated schematically in Fig. 12 using a single node temperature model in which Rcu and RTEM represent the thermal resistance of the embedded copper and the TEM respectively. In Fig. 12 the heat transfer to the solar tube resulting from the Sunlamp's 804.1 W/m2 irradiance is calculate with respect to the outer area of the cylindrical solar tube. The solar tube's ability to absorb the heat transfer from the Sunlamp is measured in terms of

Fig. 9. Spectrum irradiance measurement of the Sunlamp.


683

Fig. 12. Thermal network of the STEG.

qtube which is calculated from the temperature increase inside the tube such that,

qtube ¼ m$cp

DT Dt

(21)

in which m is the mass of the fluid inside the tube and cp is that fluid's specific heat. In order to calculate qtube, the average of the slopes from the water temperature readings illustrated in Fig. 11 are used. Furthermore, the difference in qlamp and qtube yield qloss. The results are presented in Table 1. The results presented in Table 1 show that the solar tube under the working conditions of the present study absorbs 53% of the heat transferred to it by the Sunlamp's irradiation. 4.5. Thermoelectric differential and current The thermoelectric characteristics of the device are measured by recording the electric potential in Volts generated by the thermal differential and the resultant electric current in Amperes. The results are presented in Fig. 13 showing the electric characteristics as power, voltage and current evolution curves with respect to an increasing electrical load resistance. It is shown that the electric potential increases with increasing electrical load and that, conversely the electric current decreases with increasing electrical load. Of notable importance is that the thermoelectric power output attains a peak value for an optimal electrical load. 4.6. Peak thermoelectric power production Fig. 10. Irradiance of Sunlamp at grid sensory points.

The peak thermoelectric power of the test apparatus is measured with respect to a varying electrical load on the circuit as illustrated in Fig. 13. The experimental results are repeatable due to the controlled heat emitted from the Sunlamp and the controlled heat sink into the ambient air by way of the CPU heat diffuser. In an effort to verify this, the results are measured twice generating tests results referred to as Test A and Test B. These are presented in Table 2 showing that peak power is attained upon approximate load matching as is predicted in the afore discussed electromagnetic theory for thermoelectric devices. The area of the apparatus exposed to the Sunlamp's luminance is measured to be that of the area of the reflector at 0.0897 m2 resulting in a device maximum power output per unit area of 4.82 W/m2. Considering that the available energy flux emitted by the Sunlamp is measured to be 804.1 W/m2, the peak Solar energy to electrical power conversion efficiency of the apparatus is effectively 0.60%. This conversion efficiency result is approximately 10

Table 1 Heat transfer rates.

Fig. 11. Water temperature readings in the solar tube when operating with the natural Sun and with the Sunlamp.

qlamp (W)

qtube (W)

qloss (W)

71.1

37.8

33.3

684


Fig. 13. Thermal differential converted to electrical power, electrical potential and electrical current with respect to an increasing electrical load.

fold greater than that previously reported by Lesage et al. [40] and is of the same order as that reported by He et al. [42]. The novelty in the present results lies in the fact that no external liquid pump is required to force convective cooling to the heat sink eliminating the adverse pumping penalty present in the results of He et al. [42]. 4.7. Optimal thermoelectric characteristics The normalized evolution curves with respect to electrical load resistance of the electric power, electric potential and electric current presented in Eqs. (16)e(19) are compared with the experimental data. To this end, the internal electrical resistance of the thermoelectric module is measured as detailed in Eq. (20) and presented in Table 2. The normalized evolution curves are comparing with the measured values from Test A and are illustrated in Fig. 14 showing agreement to be within a tolerance of 5% with the theoretical predictions. The results illustrated in Fig. 14 confirm that the measurements of this study are taken accurately. Furthermore, the normalized electric current versus the normalized voltage are compared with the theoretical prediction of Eq. (19). The results illustrated in Fig. 15 show agreement for both Test A and Test B with a divergence from the predicted curve for normalized electric current values greater than 1.5. The importance in the thermoelectric optimal characteristic results illustrated in Figs. 14 and 15, is that the normalized curves of Eqs. (16)e(19) apply to the present work's solar thermoelectric apparatus. This effectively means that a power point tracking systems for the optimisation of a practical application of solar thermoelectric power production can rely on Eqs. (16)e(19) for pinpointing the conditions in which maximum power is obtained.

Fig. 14. Test A: measured normalized power, normalized voltage and normalized electric current evolution curves with respect to an increasing normalized electrical load resistance compared with their respective theoretical predictions.

energy to electrical energy. Under the previously discussed operating conditions, the apparatus is exposed to the Sunlamp for a total of 240 min. From the previous calculation of the heat transfer rate of the sunlamp to the solar tube and the measured electrical power output over 240 min, Fig. 16 illustrates that the power output increases over this period reaching a maximum conversion efficiency of 0.63% relative to qlamp and a maximum conversion efficiency of 1.2% relative to qtube. This implies that the TEM is converting 1.2% of the thermal energy available to it into electrical energy. Fig. 16 illustrates that the current state-of-the-art in thermoelectric modules does not effectively exploit the thermal energy provided to it by the solar tube in the STEG application. This is seen by the apparent equality in the qtube and qout curves of Fig. 16. Therefore, improvements in thermoelectric materials would increase the overall performance of the present work's STEG apparatus. For example, if the thermoelectric material were to provide a 20% conversion efficiency, the present apparatus would have an overall system efficiency of 10.6% since the solar tube converts 53% of the available solar energy to heat. 4.9. Field test application In this study, the test apparatus is applied to the solar and wind conditions of the natural environment of the Yucatan peninsula in

4.8. Thermal transport and thermoelectric power As illustrated in Fig. 12, the two main energy losses in the thermal transport mechanism is due to the solar tube's ability to absorb solar radiation and the TEM's ability to convert thermal

Table 2 Peak power output and electrical load characteristics of solar thermoelectric device.

Test A Test B

Pmax (W)

RL (U)

Ri (U)

Voc (V)

VL (V)

I (A)

0.423 0.432

5.519 5.622

5.263 5.146

2.985 2.985

1.528 1.558

0.277 0.277

Fig. 15. Normalized electric current versus normalized electric potential for Test A and Test B compared with their theoretical evolution curve.


Fig. 16. Thermal transport and power output of the STEG over time.

Quintana Roo, Mexico. This region is considered an ideal location for a first field test of the presented technology due to its favourable wind and solar energy resources [e.g., Refs. [56,57]]. Furthermore, this region presents a potential application for the tested apparatus since more than 5% of Quintana Roo's residents live in dwellings that are isolated from the local electricity grid. Indeed, the large number of off-grid homes in the State of Quintana Roo is attributed to the high costs of electricity [58]. Since home-installed photovoltaic panels and wind turbines are less economic than local combustion generators, many isolated homes invest in combustion generators resulting in adverse noise and pollution. The Qunitana Roo local conditions therefore provide a motivation for advancing the state-of-the-art in STEG technology. The test apparatus is exposed to the natural sun on the Unigica de Cancún, Quintana Roo campus for versidad Tecnolo approximately 60 min starting a 9:42 local time. During this period, the average wind velocity is measured to be 2.02 m/s using a Anaheim H400 anemometer. The hot and cold side temperatures of the TEM are measured using a Fluke 87 Series V with type-K thermocouple in contact with the adjacent material. The preliminary results illustrated in Fig. 17 represent the module's internal electromotive force generated over the test period. 5. Conclusion The design and assembly of an apparatus which effectively converts thermoelectrically Solar energy to electrical power is presented. The main thesis of this work is to investigate the

Fig. 17. Open circuit voltage measurement of the STEG apparatus during a field test application.

685

management of a thermal dipole in which the heat source is solar radiation and the heat sink is Wind aided thermal diffusion into the ambient air for the purpose of mobilising charge carriers in a thermoelectric semiconductor material. This work presents novel results since the energy input to the test apparatus simulates naturally occurring Solar and Wind energy and requires no external power sources such as fluid pumps to generate the generated thermoelectric power. Optimal thermoelectric characteristics are derived and compared with the validating data and preliminary results of a field test application are provided. The results advance the state-of-the-art in solar thermoelectric power production by enhancing the thermal transport mechanism of the double walled vacuum solar tube to a commercially available thermoelectric module. More specifically, a single 30 30 3.75 mm thermoelectric module attains a maximum power output of 432 mW under operating conditions which simulate local Solar radiance conditions at the 45 280 3800 N parallel of latitude and 50 m above sea level. The results show that the device effectively converts 0.60% of the available 804.1 W/m2 to electrical power resulting in 4.82 W/m2. Furthermore, theoretical predictions of peak power operating conditions for a solar thermoelectric generator (STEG) are presented. In conclusion, the present work demonstrates the potential to convert Solar and Wind energy by exploiting the thermoelectric phenomenon without the use of photovoltaic materials nor turbines and without any adverse pumping penalty. Acknowledgements The authors gratefully acknowledge the generous financial gep de l'Outaouais and the equipsupport of the Fondation du Ce ment support of Hydromax Gatineau. The authors sincerely thank Dr. Karin Hinzer and the University of Ottawa's SunLab for their collaborative support throughout this project. References [1] D. Kraemer, L. Hu, A. Muto, X. Chen, G. Chen, M. Chiesa, Photovoltaicethermoelectric hybrid systems: a general optimization methodology, Appl. Phys. Lett. 92 (24) (2008) 243503. [2] W.V. Sark, Feasibility of photovoltaicethermoelectric hybrid modules, Appl. Energy 88 (8) (2011) 2785e2790. [3] G. Rockendorf, R. Sillmann, L. Podlowski, B. Litzenburger, PV-hybrid and thermoelectric collectors, Sol. Energy 67 (4) (1999) 227e237. rez-Garcıa, P. Vorobiev, U. Dehesa[4] R. Zakharchenko, L. Licea-Jimenez, S.A. Pe Carrasco, J.F. Perez-Robles, Y. Vorobiev, Photovoltaic solar panel for a hybrid PV/thermal system, Sol. Energy Mater. Sol. Cells 82 (1) (2004) 253e261. vez-Urbiola, Y.V. Vorobiev, L.P. Bulat, Solar hybrid systems with [5] E.A. Cha thermoelectric generators, Sol. Energy 86 (1) (2012) 369e378. [6] X. Ju, Z. Wang, G. Flamant, P. Li, W. Zhao, Numerical analysis and optimization of a spectrum splitting concentration photovoltaicethermoelectric hybrid system, Sol. Energy 86 (6) (2012) 1941e1954. [7] Y. Deng, W. Zhu, Y. Wang, Y. Shi, Enhanced performance of solar-driven photovoltaicethermoelectric hybrid system in an integrated design, Sol. Energy 88 (2013) 182e191. [8] I.I. Kokhova, Hu.N. Malevskii, A.I. Tsvetkov, Estimating capacity of solar thermoelectric generator (STEG) panels, Appl. Sol. Energy 15 (6) (1979) 20e25. [9] K.P. Suleebka, High temperature solar thermoelectric generator, Appl. Energy 5 (1) (1979) 53e59. [10] R. Amatya, R.J. Ram, Solar thermoelectric generator for micropower applications, J. Electron. Mater. 39 (9) (2010) 1735e1740. [11] K. McEnaney, D. Kraemer, Z. Ren, G. Chen, Modeling of concentrating solar thermoelectric generators, J. Appl. Phys. 110 (7) (2011), 074502-1e074502-6. [12] G. Chen, Theoretical efficiency of solar thermoelectric energy generators, J. Appl. Phys. 109 (2011) 104908. [13] D. Kraemer, K. McEnaney, M. Chiesa, G. Chen, Modeling and optimization of solar thermoelectric generators for terrestrial applications, Sol. Energy 86 (5) (2012) 1338e1350. [14] F. Demichelis, E. Minetti-Mezzetti, M. Agnello, E. Tresso, Evaluation of thermophotovoltaic conversion efficiency, J. Appl. Phys. 53 (12) (1982) 9098e9104. [15] C. Wu, Analysis of waste-heat thermoelectric power generators, Appl. Therm. Eng. 16 (1) (1996) 63e69.

686


[16] J.J. Bonilla, J.M. Blanco, L. Lopez, J.M. Sala, Technological recovery potential of waste heat in the industry of the Basque Country, Appl. Therm. Eng. 17 (3) (1997) 283e288. [17] T. Li, G. Tang, G. Gong, G. Zhang, N. Li, L. Zhang, Investigation of prototype thermoelectric domestic-ventilator, Appl. Therm. Eng. 29 (10) (2009) 2016e2021. -Potvin, Experimental analysis of peak power output of a [18] F.J. Lesage, N. Page thermoelectric liquid-to-liquid generator under an increasing electrical load resistance, Energy Convers. Manag. 66 (2013) 98e105. [19] F.J. Lesage, E.V. Sempels, N. Lalande-Bertrand, A study on heat transfer enhancement using flow channel inserts for thermoelectric power generation, Energy Convers. Manag. 75 (2013) 532e541. ~o, E.V. Sempels, F.J. Lesage, Net thermoelectric generator [20] C. Amaral, C. Branda power output using inner channel geometries with alternating flow impeding panels, Appl. Therm. Eng. (2014), http://dx.doi.org/10.1016/j.applthermaleng. 2013.12.044. [21] H.J. Goldsmid, J.E. Giutronich, M.M. Kaila, Solar thermoelectric generation using bismuth telluride alloys, Sol. Energy 24 (5) (1980) 435e440. [22] D.M. Rowe, A high performance solar powered thermoelectric generator, Appl. Energy 8 (4) (1981) 269e273. [23] J. Chen, Thermodynamic analysis of a solar-driven thermoelectric generator, J. Appl. Phys. 79 (5) (1996) 2717e2721. [24] B. Lenoir, A. Dauscher, P. Poinas, H. Scherrer, L. Vikhor, Electrical performance of skutterudites solar thermoelectric generators, Appl. Therm. Eng. 23 (11) (2003) 1407e1415. [25] C.C. Bomberger, P.M. Attia, A.K. Prasad, J.M. Zide, Modeling passive power generation in a temporally-varying temperature environment via thermoelectrics, Appl. Therm. Eng. 56 (2013) 152e158. [26] J.Y. Jang, Y.C. Tsai, Optimization of thermoelectric generator module spacing and spreader thickness used in a waste heat recovery system, Appl. Therm. Eng. 51 (2012) 677e689. [27] L.A. Weinstein, K. McEnaney, G. Chen, Modeling of thin-film solar thermoelectric generators, J. Appl. Phys. 113 (16) (2013), 164504-1e164504-8. [28] X.W. Wang, H. Lee, Y.C. Lan, G.H. Zhu, G. Joshi, D.Z. Wang, Z.F. Ren, Enhanced thermoelectric figure of merit in nanostructured n-type silicon germanium bulk alloy, Appl. Phys. Lett. 93 (19) (2008), 193121e193121. [29] A. Inagoya, D. Sawaki, Y. Horiuchi, S. Urata, R. Funahashi, I. Terasaki, Thermoelectric module made of perovskite cobalt oxides with large thermopower, J. Appl. Phys. 110 (12) (2011), 123712e123712. [30] A. Bhardwaj, D.K. Misra, J.J. Pulikkotil, S. Auluck, A. Dhar, R.C. Budhani, Implications of nanostructuring on the thermoelectric properties in half-Heusler alloys, Appl. Phys. Lett. 101 (13) (2012), 133103e133103. [31] J. Xiao, T. Yang, P. Li, P. Zhai, Q. Zhang, Thermal design and management for performance optimization of solar thermoelectric generator, Appl. Energy 93 (2012) 33e38. [32] A.S. Al-Merbati, B.S. Yilbas, A.Z. Sahin, Thermodynamics and thermal stress analysis of thermoelectric power generator: influence of pin geometry on device performance, Appl. Therm. Eng. 50 (2012) 683e692. [33] H. Ali, A.Z. Sahin, B.S. Yilbas, Thermodynamic analysis of a thermoelectric power generator in relation to geometric configuration device pins, Energy Convers. Manage. 78 (2014) 634e640. [34] N. Vatcharasathien, J. Hirunlabh, J. Khedari, M. Daguenet, Design and analysis of solar thermoelectric power generation system, Int. J. Sustain. Energy 24 (3) (2005) 115e127. [35] C.A. Mgbemene, J. Duffy, S. Hongwei, S.O. Onyegegbu, Electricity generation from a compound parabolic concentrator coupled to a thermoelectric module, J. Sol. Energy Eng. 132 (3) (2010). [36] Y. Yang, Q. Yan, R. Zhai, A. Kouzani, E. Hu, An efficient way to use medium-orlow temperature solar heat for power generationeintegration into conventional power plant, Appl. Therm. Eng. 31 (2) (2011) 157e162.

[37] X.F. Zheng, Y.Y. Yan, K. Simpson, A potential candidate for the sustainable and reliable domestic energy generation e thermoelectric cogeneration system, Appl. Therm. Eng. 53 (2012) 305e311. [38] K. Yazawa, A. Shakouri, System optimization of hot water concentrated solar thermoelectric generation, in: 2010 3rd International Conference on Thermal Issues in Emerging Technologies Theory and Applications (ThETA), IEEE, December 2010, pp. 283e290. [39] K. Yazawa, V.K. Wong, A. Shakouri, Thermal challenges on solar concentrated thermoelectric CHP systems, in: 2012 13th IEEE Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic Systems (ITherm), IEEE, May 2012, pp. 1144e1150. [40] F.J. Lesage, R. Pelletier, L. Fournier, E.V. Sempels, Optimal electrical load for peak power of a thermoelectric module with a solar electric application, Energy Convers. Manag. 74 (2013) 51e59. [41] W. He, Y. Su, S.B. Riffat, J. Hou, J. Ji, Parametrical analysis of the design and performance of a solar heat pipe thermoelectric generator unit, Appl. Energy 88 (12) (2011) 5083e5089. [42] W. He, Y. Su, Y.Q. Wang, S.B. Riffat, J. Ji, A study on incorporation of thermoelectric modules with evacuated-tube heat-pipe solar collectors, Renew. Energy 37 (2012) 142e149. [43] Z.N. He, H.C. Ge, F.L. Jiang, W. Li, A comparison of optical performance between evacuated collector tubes with flat and semi-cylindric absorbers, Sol. Energy 60 (2) (1997) 109e117. [44] L.J. Shah, S. Furbo, Theoretical flow investigations of an all glass evacuated tubular collector, Sol. Energy 81 (2007) 822e828. [45] L. Ma, Z. Lu, J. Zhang, R. Liang, Thermal performance analysis of the glass evacuated tube solar collector with U-tube, Build. Environ. 45 (9) (2010) 1959e1967. [46] M. Zhang, L. Miao, Y.P. Kang, S. Tanemura, C.A. Fisher, G. Xu, G.Z. Fan, Efficient, low-cost solar thermoelectric cogenerators comprising evacuated tubular solar collectors and thermoelectric modules, Appl. Energy 109 (2013) 51e59. [47] S. Karabetoglu, A. Sisman, Z. Fatih Ozturk, T. Sahin, Characterization of a thermoelectric generator at low temperatures, Energy Convers. Manag. 62 (2012) 47e50. [48] R.Y. Kim, J.S. Lai, A seamless mode transfer maximum power point tracking controller for thermoelectric generator applications, IEEE Trans. Power Electron. 23 (5) (2008) 2310e2318. [49] R.Y. Kim, J.S. Lai, B. York, A. Koran, Analysis and design of maximum power point tracking scheme for thermoelectric battery energy storage system, IEEE Trans. Ind. Electron. 56 (9) (2009) 3709e3716. [50] X. Zhang, K.T. Chau, An automotive thermoelectricephotovoltaic hybrid energy system using maximum power point tracking, Energy Convers. Manag. 52 (1) (2011) 641e647. [51] D.M. Rowe, Thermoelectrics Handbook Macro to Nano, Taylor & Francis Group, 2006. [52] D. Kraemer, B. Poudel, H.P. Feng, J.C. Caylor, B. Yu, X. Yan, G. Chen, Highperformance flat-panel solar thermoelectric generators with high thermal concentration, Nat. Mater. 10 (7) (2011) 532e538. [53] M. Hodes, On one-dimensional analysis of thermoelectric modules (TEMs), IEEE Trans. Compon. Packag. Technol. 28 (2005) 218e229. [54] J. Eakburanawat, I. Boonyaroonate, Development of a thermoelectric battery charger with microcontroller-based maximum power point tracking technique, Appl. Energy 83 (2006) 687e704. [55] C.A. Gueymard, D. Myers, K. Emery, Proposed reference irradiance spectra for solar energy systems testing, Sol. Energy 73 (6) (2002) 443e467. [56] National Renewable Energy Laboratory (NREL), International Wind Resource Maps 8-Dec-2000, From http://www.nrel.gov/wind/images/mexico-yucatanwindmap.gif. ~ a, Hermilio O. Ortega Navarro, Prospectiva de Energías [57] Julio A. Valle Peren Renovables 2012e2016, 2012. [58] Instituto Nacional Estadística y Geografía, Mexico en cifras, 2010, From http:// www3.inegi.org.mx/sistemas/mexicocifras/default.aspx?e¼23.