~ ) Pergamon 0196-8904(94)00038-7 Energy Convers. Mgmt Vol. 36, No. I, pp. 41 59, 1995

Copyright ~ 1995 Elsevier Science Ltd Printed in Great Britain. All rights reserved

0196-8904/95 $9.50 + 0.00

SIMULATION AND OPTIMIZATION OF ENERGY SYSTEMS FOR IN-BIN DRYING OF CANOLA GRAIN

(RAPESEED)

G. J. SCHOENAU, E. A. ARINZE and S. SOKHANSANJ College of Engineering, University of Saskatchewan, Saskatoon, Saskatchewan, Canada S7N 0W0

:Received 8 October 1993; received for publication 7 September 1994)

A ~ t r a e t - - E n e r g y utilization systems optimization and management strategies for in-bin drying of canola were investigated by using a validated computer simulation model and typical weather data for a prairie location in North America. The use o f different energy systems, including natural gas, propane, electricity, solar energy, and combined natural gas and solar energy for drying grain within 15 days with airflow rates of 0.5-2 m3/min t, initial grain moisture contents of 13, 16 and 19%, and three harvest dates in August, September and October, was simulated for 10% and 8% moisture contents average-dry and through-dry policies. The drying systems were optimized by considering the total annual cost of a drying system within set bounds of drying time (~< 15 days) and spoilage index (SI < 1.0). Cont inuous fan operation with 1.5-2 m3/min t ambient air with about 9-26 MJ/t fan energy consumption was required to dry canola grain to 10% and 8% average-dry and through-dry moisture contents in 15 days or less August at 19% initial moisture content or less. Supplemental heat, by raising the ambient temperature by 5-10°C, maintaining the plenum temperature at 20°C and solar heating, must be applied to successfully dry the product in September and October. Solar heating for drying was found to be more cost effective than other supplemental heat systems provided a well designed flat-plate solar collector for air heating can be found for use in locations with good solar energy availability. Heating the drying air with natural gas or propane was the cost effective for situations where the use of conventional energy systems is preferable to renewable energy sources in grain drying operation.

Energy systems utilization Grain drying Management Optimization Simulation

N O M E N C L A T U R E

A = Floor area (m s) a = Material constant in equation (1) C = Specific heat (J/kg°C)

C~ ,C2,C3 = Material constants in equations (3a) and (3b) C, = Electrical energy cost ($) C r = Fuel energy cost ($) D = Time in storage before grain loses 5% germination (d)

dx = Thickness of grain layer (m) E e = Electrical energy (k J) Ef = Fuel energy (k J)

FE = Fan and motor efficiency (fraction) FP = Fan power (W)

Fr = Ratio of annual capital cost to initial cost of equipment (fraction) f = Factor representing incidence of bad weather fraction of fines in grain (fraction)

h = Number of operating hours in year (h) i = Interest rate (decimal)

L = Change in heat of vaporization or bin height (J/kg or m) M = Moisture content (% or decimal)

M R = Moisture ratio (dimensionless) M r = Equilibrium moisture content (% dry basis) M 0 = Initial moisture content (% dry basis) N = Material constant in equation (3a) n = Years of useful life (a)

P~ = Value of grain per tonne ($) Q = Airflow rate (m3/s m 2 or m3/min t) r = Resale value as fraction of initial cost (fraction)

R c = Repair cost as fraction of initial cost (fraction) RH = Relative humidity (%. decimal)

41

42 SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN

Sl = Spoilage or storage index (dimensionless) T = Temperature (°C)

AT = Temperature increase to ambient air (°C) T~ = Air temperature (°C) T~ = Grain temperature (°C) t = Time (s, h)

At = Time step or increment (s, h) 14" = Mass of grain at initial moisture content (t) p = Product density (kg/m 3)

Subscripts

a = Air, ambient e = Equilibrium f = Final i = Initial, inlet 0 = Initial p = Product, plenum v = Vapor w = Water

I N T R O D U C T I O N

Canola is a major oilseed crop grown in many parts of the world, including Canada, Europe (especially U.K., Sweden, France, Germany and Italy), Australia, New Zealand, China and India. Canola production is well adapted to cool moist weather. The current average world production ofcanola is over 20 million tonnes, worth about 5 billion U.S. dollars, and Canada produces about 25% of the world total [1]. Canola contains over 40% oil and 20% protein. The canola oil is extracted from the seeds for production of edible oil, margarine, and salad dressings in the food industry, and the by-product, which is rich in protein, is used as a high protein meal for feed in livestock production. Because of its oil content, canola is extremely susceptible to spoilage if not properly dried before or during storage. With the increasing supply of canola as a source of edible oil, the quantity dried and stored, and the time for which it may remain in store are also increasing.

For marketing, canola is considered dry at 10-13% moisture content. Unless otherwise specified, all moisture contents are on the wet basis. However, about 8% moisture content at temperatures below 20°C are necessary for safe storage [2]. In most producing areas, outdoor weather conditions during harvest are not favorable for complete field drying of canola to safe storage conditions. In addition, shattering losses and mechanical damage during combining are lower if the product is harvested at a higher moisture content than the safe storage condition. Harvesting canola at a higher moisture content also allows a longer harvest season, earlier harvesting and reduced field losses. All the above factors necessitate artificial drying of canola after harvest for safe storage and marketing.

The type of equipment used in any artificial drying process depends largely on mass and energy transfer between the drying air and the product. The most common method of drying agricultural products is forced air circulation through the product held in a suitable container, such as bins, silos or chambers. In-bin drying of canola is becoming a required practice in most producing countries, especially in the Canadian Prairies, which are the major canola producing areas in North America. In natural or unheated air systems, a fan continuously forces ambient air through the grain mass in the bin until the grain is dried to a safe storage moisture content. The drying rate depends on the initial moisture content and temperature of the product, air circulation rate, entering conditions of the circulated air as given by dry-bulb temperature and relative humidity, length of flow path through the product and elapsed time since the beginning of the drying operation. Due to frequent incidences of unfavorable ambient air drying conditions, the outdoor air is preheated before circulation through the product to promote faster drying and prevent product spoilage.

Preheating the drying air increases the heat transfer rate to the product and, consequently, the product temperature and vapor pressure [3, 4]. Preheating of the air, therefore, raises the drying potential of the air and decreases the total time required for drying the product from given initial and required final safe moisture contents. Depending on the initial moisture content and

SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN 43

environmental conditions, the drying of canola must be completed fast enough to prevent rapid product deterioration by microorganisms and enzymes. However, in addition to capital and operating costs, the heated air drying could potentially promote grain spoilage or quality deterioration during drying and over-drying if drying is not managed properly [5].

Conventional energy sources, such as natural gas, propane, fuel oils (diesel and kerosene) and electricity, and renewable energy sources, such as solar energy and biogas, have been utilized for heating the drying air in various crop drying systems. While availability and costs influence the selection of any of these energy resources for drying in a given location, the important management factors to be considered in the operation of a dryer include drying equipment costs, specific energy consumption and cost, drying time and drying cost and degree of product spoilage during drying. The main objective of this study is to simulate and optimize energy use systems for in-bin drying of canola grain under typical weather conditions in the Canadian Prairies. The optimal operation of existing driers is the main focus. The energy systems considered in this study include unheated air, heated air with natural gas, heated air with propane, heated air with electricity, heated air with solar energy, and heated air with solar energy and natural gas (hybrid). The optimal system for drying grain with unheated or heated air is the one that gives a low drying cost but still dries the grain to a moisture content safe for storage without any spoilage due to mold growth or enzymes.

E N E R G Y U S E S Y S T E M S IN D R Y I N G

Natural gas and propane indirect air heaters

Propane has a lower heating value of about 89,040 kJ/m 3 and, for natural gas, the usual range is 37,300-39,200 kJ/m 3 with a mean value of 38,250 kJ/m 3 [6]. A typical natural gas or propane furnace consists of (i) the cabinet or casing, (ii) heat exchangers, (iii) the combustion system including burners and controls, (iv) a draft hood or flue collector box, (v) a circulating air blower or fan and electric motor or engine, (vi) an air filter (where necessary) and (vii) other accessories such as a dehumidifier.

Electric air heater

In an electric air heater or furnace, the resistance-type heating elements either heat the circulating drying air directly or through a metal sheet enclosing the resistance element. Electric powered air heaters have various configurations. In a typical arrangement for electric forced air heat, air enters the downstream or bottom of the furnace and passes through the filter, then flows into the blower.

Flat-plate solar air heaters

The solar heating function in forced circulation crop drying generally consists of separate flat-plate collectors to preheat the ambient air and reduce its relative humidity before passing the air to the drying bin. Axial-flow or centrifugal fans are used to circulate the ambient air through the solar collector and product to be dried. In the design of solar collectors for heating drying air, the collector size for a given airflow rate is determined by considering the maximum allowable range of air temperatures for drying various crops [7, 8].

Natural or unheated air

A natural or unheated forced air drying system consists of a bin for storage of the grain, a fan and electric motor or engine to force the air through the product and an appropriate duct system and perforated floor for uniform air distribution through the product. The power required to operate the fan becomes quite large for greater grain depths and is generally considered uneconomical for drying. There is usually a change in the temperature of the ambient air due to waste heat from the fan. This relatively small amount of air temperature rise of about 2 ~ ° C [9] has the effect of significantly increasing the drying air potential or drying rate.

S IMULATION MODEL

A computer program was developed to simulate in-bin drying of canola grain and predict grain spoilage under the simulated conditions. The drying model, which is a modified version of a

44 SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN

near-equilibrium model [4], simulates heat and moisture transfer in the grain bulk in the bin and determines the time required to dry the entire product from a given initial moisture content to a desired final moisture content. The spoilage model predicts the storability of grain mass by correlating the storage life of the grain with the grain temperature and moisture content. Capital investment and operating costs associated with various drying schemes are needed for the economic analysis.

Drying model

The drying rates of agricultural grains are commonly calculated using a thin layer drying equation. A modified equation for canola, derived from experimental work [10], is:

where

MR = moisture ratio

(M - Me)

(M0 - M e ) = M R = e -~'" (1)

where

RH = equilibrium relative humidity (decimal) T = air or grain temperature (°C)

Me = equilibrium moisture content (% dry basis) C~, C2 and N = constants which depend upon the material.

The experimental values of the constants in equation 3(a) for Tobin canola are [12]:

C~ =0.0005056, C2=40.1204 and N = 1.5702.

For adsorption or rewetting (which may occur especially at the top layers of the grain, depending on the drying air condition, grain moisture content and depth in the bed), the equilibrium moisture content and relative humidity relationship is given by the modified Halsey equation [13, 14] as:

RH = exp[(-exp(C~ - C 2 T ) M ~ c3] (3b)

where the constants C~, (72, C3 are 3.415, 0.01191 and 1.820, respectively [14]. The thin layer drying simulation method for a deep-bed of material adopted in this study follows

the experimentally validated thin layer simulation procedure [4] in which the average changes in moisture content and temperature on a thin layer of material are calculated over a discrete time interval At. The development of the modified model is presented in detail in Refs [1, 5, 15].

In the simulation of a deep-bed batch drying process, the bed height h is divided into a number of thin layers of thickness dx each. Drying is achieved by continuously passing heated or unheated

M0 = initial moisture content of material (% of dry basis) M, = equilibrium moisture content of material (% dry basis) M = moisture content (% dry basis) of material at time t k = drying constant, dependent on the materials (s-t or h - t ) t = time (s or h) a = material constant.

An average value of a = 0.818 for canola [10]. A regression of k for canola with air temperature and relative humidity is [10]:

k = -0 .0257 + 0.00094T a + 0.00031RH (2)

where

Ta = air temperature (°C) RH = relative humidity (%).

The desorption or drying equilibrium moisture content and relative humidity relationship, described in Refs [11, 12], in the form of a modified Henderson equation [13] for canola is:

RH = 1 - e x p [ - C, (T + C2)M~] (3a)

SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN 45

air through the static bed of material in one direction, usually from the bottom to the top section. From the first thin layer at the bottom section, the drying air evaporates moisture from the material and carries it to the next thin layer. As the air absorbs moisture, its temperature decreases and its ability to pick up more moisture (drying potential) decreases. A deep-bed consisting of a number of thin layers is, therefore, simulated by calculating the air and moisture changes as the drying air passes from one thin layer of material to the next layer. Each layer dries to equilibrium conditions for a short time interval, and the exhaust air from one layer is used as the input drying air to the next. The drying procedure continues for a number of drying time intervals until the desired final moisture content of the material is achieved.

One of the important input parameters used in the simulation is the specific heat of the product Co, which is calculated in terms of the grain moisture content M and temperature Tg as [16]:

Cd = 1270 + 34M -- 1.33Tg for Tg ~< 0°C and M >~ 10% (4a)

C d = 1 2 6 5 + 3 0 M + 5 . 9 5 T g for Tg>0°C and M > 1 % . (4b)

The calculated specific heat for the product is expressed in J/kg °C. The values of the other assumed constants are: C, = 1008, Cv = 2000, Cw = 4180, hfg = 22,50,000, and p = 700 where C,,, Cv and Cw are the specific heat of air, vapor and water (J/kg °C), respectively, hrg is the heat of vaporization of water (J/kg) and p = density of the Tobin variety canola (kg/m 3). The psychromet- ric relations used in the simulation were obtained from Refs [6, 17].

Resistance of bulk canola to airflow

The following equation has been developed from experimental data for calculation of the resistance of canola to airflow [18]:

AP 5.22 × 104Q 2 (1 + 1.75f) (5)

L ln(1 + 7.27Q )

where AP is the resistance of the grain to airflow in Pa, L is grain depth, Q is the airflow rate in m3/s per square meter of the floor area a n d f i s the fraction of fines in the grain in decimal. If the grain is clean, then f = 0. Equation (5) is a standard equation, used by the designers of low temperature drying systems, to estimate the pressure required to push a given airflow into the bin [17]. Once AP has been calculated, the fan power (FP) is estimated from:

FP = APQA/FE (6)

where

FP = fan power (W) FE = efficiency of the fan and motor (usually a value of about 0.5 is used)

A = floor area (m2).

Flat-plate solar collector dynamic performance model

At any instant and weather conditions, the collector outlet air or dryer inlet air temperature is determined following the procedure in a dynamic performance simulation model of flat-plate solar collectors, Refs [19, 20]. In the solar collector model, three ordinary differential equations are written for the collector fluid, absorber plate and transparent cover temperatures. The differential equations are solved numerically by the Runge-Kutta predictor-corrector method to obtain the collector outlet fluid temperature and other parameters. The Runge-Kutta method has been chosen because memory requirements for the computer can be minimized if an adequate or suitable time spacing which does not reduce accuracy is chosen [19].

For the numerical solution of the differential equations by the Runge-Kutta predictor-corrector method, the values of the collector dependent variables are predicted by using their values at time t = t - At and the values of the external boundary parameters at time t to obtain the derivatives and new values of the dependent variables at time t. The iteration process is stopped as soon as the values of the dependent variables calculated in two consecutive iteration steps are equal or if the difference in values of the variable in two consecutive iteration steps is within the degree of accuracy specified in the computer program. When the tolerable difference test is satisfied, the

46 SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN

solution for the time step t is completed, and the predicted values are used in order to get new solutions at time t = t + At. This procedure is repeated until the solutions for all time steps are obtained. In most computer runs, convergence was obtained in two iterative steps. An advantage of the above predictor-corrector technique is that the iterative calculations occurring during a single time step are performed at a constant value of time.

The output of a collector fluctuates greatly during the course of the day and from day to day. It is, therefore, not possible to accurately determine the collector output or performance at any given instant from previously measured meteorological data. It is, however, possible to determine the average collector output with statistically average data, which may cover long periods of time as with the typical meteorological year (TMY) data available for many locations in North America. The method used in obtaining the incident and transmitted solar radiation on tilted surfaces from the TMY horizontal total solar radiation data is presented in Refs [3, 8].

Spoilage simulation model

The spoilage model predicts the allowable storage times of grain before significant quality loss occurs to the grain. The quality loss may be signified by a drop in seed germination capacity, visible microflora, seed clumping or by other deterioration indicators. Based on germination data [21], a set of two regression equations for predicting allowable storage times for canola varieties before the germination capacity drops by 5% was developed [11]:

lOgloD=6.224-O.302M-O.O69Tg for M < l l % (Ta)

logloD=5.278-O.206M-O.O63Tg for M ~ > l l % (7b)

where

D = time in storage until 5% drop in germination (days) M = grain moisture content (% w.b.) Tg = grain temperature (°C).

Equation (7) predicts the allowable storage times when canola is stored at constant temperatures and moisture contents. During a drying process, both temperature and moisture content vary with time. To predict grain spoilage deterioration under dynamic or changing conditions, a spoilage index (SI) was used. A value of D was computed at each interval At from equation (7), the ratio At/D was calculated, and the ratios were cumulated. Theoretically, grain loses 5% of its germination when the sum of the computed At/D values for each layer over the simulated drying period equals unity:

s , 1 i = 1 i ~

where n is the number of simulated time steps. SI is a spoilage or storage index, and its instantaneous value represents the progress of grain spoilage. A spoilage index of 1 or greater indicates that the allowable storage time has elapsed, and the 5% loss in germination has occurred to the canola.

Calculation of equipment, labor, energy and over-drying costs

For the purpose of comparing different drying policies utilizing various energy systems and cost optimization, the following costs were calculated:

Equipment capital costs, CI. The ratio of the annual capital cost to the initial cost of the drying equipment Fr, was expressed as follows [22]:

. { l + r l - r ) l - r (9) F r = l ~ T + ~ + n

where i is the annual interest rate (10%), r is the ratio of the salvage value to the original value (10%) and n is the life of the drying system. For the perforated floor, n = 20 yr. For the gas or electric heater, fan and controllers, n was estimated from:

n = 5040/h (10)

SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN 47

where 5040 is the expected life in hours, and h is the number of operating hours in a year. The annual equipment capital cost Ct was calculated by:

C, = Co(rr+ Rc) (11)

where R~ is a ratio of the maintenance costs to the initial cost Co of the drying equipment (P~ = 0.01).

Labor or mixing cost, C2. Mixing costs were included if the grain was dried to an average of 10 or 8% "avg-dry" and not included if the grain was dried through to top layer at 10 or 8% "thru-dry".

Energy cost, C3. The energy cost consisted of both electrical energy and fuel energy. The total energy cost was calculated as follows:

C3 = EeCe + EfCr (12)

where Ee and Er are the equivalent annual electricity and fuel consumption. Ce is the unit cost of electricity and Cr is the unit cost of fuel (natural gas or propane).

Cost of over-drying, C4. Over-drying in " thru-dry" can represent an economic loss at the time of grain sale. Number 1 Canadian canola can be sold on a mass basis at a maximum moisture content of 10% without penalty. The following equation was used to estimate the cost of over-drying:

C 4 = W ( 1 - M ' / I O 0 ) ( 1090 10 ~ r ) p , M f (13)

where M~ and Mf are the percent average wet and dry moisture contents, respectively. P~ is the price of number 1 canola. W is the total mass of wet canola in the bin dried in a year.

Input data for the simulation program Using the models presented above, a computer program was developed to simulate in-bin drying

and spoilage or grain quality change during drying of canola for various inlet air drying conditions, including natural air, gas-heated, electric heated and solar heated air drying. The simulated bin had a perforated floor and was divided into a number of vertical layers of 0.1 m thick each.

The simulated drying conditions were based on harvest conditions and drying practices common to the Canadian Prairies. The input ambient conditions (air temperature and relative humidity, and solar radiation) to the simulation were Typical Meteorological Year (TMY) data for Saskatoon in Western Canada developed by the University of Waterloo, Solar Energy Laboratory. The TMY file is constructed from a modal analysis of actual hourly weather data for at least 30 y. The TMY file contained hourly dry bulb temperature, relative humidity, horizontal total solar radiation, wind speed and ground global solar reflectance.

For the natural air drying condition, the only change in the air condition before it enters the plenum is the air temperature rise (AT) due to fan waste heat. The fan waste heat was assumed to be 50% of the total fan energy requirement. The fan power was obtained as a function of the pressure drop through the bin drying system and airflow rate from equations (5) and (6).

For the gas-heated and electric heated drying conditions, the ambient air flowing through the fan passes through a gas or electric heater with or without humidistat control, where the air is tempered to the desired temperature and humidity before it enters the plenum. The heating values of natural gas and propane were 38.25 and 89.04 MJ/m 3, respectively [6].

In the case of solar-heated drying, the ambient air is first passed through a flat-plate solar collector, where the air is preheated before passing through the fan and perforated floor plenum to the bin. The flat-plate solar collector dynamic performance model was used as a subroutine in the main simulation model to calculate the collector output air temperature. The collector output air temperature changes with solar radiation intensity during the day. The solar collector area used was 1.2 m 2 per m3/min airflow [23, 24]. The collector slope was also equal to the latitude angle (52.1TN) of the location.

The grain was Tobin canola variety with initial moisture contents of 13, 16 and 19%, and a bulk density of 700 kg dry-matter/m 3. The bin diameter was 4.3 m and the grain bed depth of 4.3 m was ventilated by a centrifugal electric fan at various air flow rates of 0.5, 1, !.5 and 2 m3/min t. Drying

E C M 36:1 I )

48 SCHOENAU e t al.: IN-BIN DRYING OF CANOLA GRAIN

was assumed to start at zero hour on 1 August, 1 September and 1 October. These are typical harvest months in the Prairies. Grade 1 canola cost of $227/t was obtained from the Pioneer Grains Company at Saskatoon. Electricity, natural gas and propane costs of $0.518/kWh, $0.127/m 3 and $0.322/1 were also obtained from the SaskPower and SaskEnergey Corporations at Saskatoon. From several other market sources, the estimated costs for the bin floor, gas heater, humidistat control and thermostat were $800, $1200 and $175, respectively. The fan costs were $660, $1000, $1275 and $1520 for 0.5, 1, 1.5 and 2 m3/min t airflow capacities, respectively. The costs of the frame or panel and top cover (polyethylene film) of the solar collector were $17/m 2 and $0.8/m', respectively. The labor cost for mixing and stirring grain in the bin was $30 per batch.

Drying schemes and policies (average-dry and through-dry) The two drying policies investigated are the "avg-dry" and "thru-dry". In the "avg-dry", the

desired final average moisture of the grain is attained by mixing or stirring the grain bulk after drying. This involves extra investment in stirring equipment and labor charges for mixing after drying and more frequent inspection during drying. This policy may lead to possible under-drying and grain spoilage later, as it is difficult to ascertain accurately when the bottom and top layers of the bin have dried to such moisture levels necessary to attain the desired final average moisture content. Experience, therefore, in determining the safe point to stop drying in this case is crucial. In the "thru-dry", the drying is continued until the top layer of the bin has reached the desired final moisture content and, consequently, the lower grain layers are usually over-dried. Although the weight loss due to over-drying is a potential cost to the farmer, the probability of grain spoilage in storage after drying is relatively low in this case, since all grain layers in the bin are initially dried to a safe moisture level. Both policies, which are commonly practiced, therefore, have merits and demerits.

In each drying simulation, drying continued until either the average grain moisture content reached 10 or 8% in the "avg-dry" case or the top layer grain moisture content reached 10 or 8% for the "thru-dry" case. The drying simulation outputs were average moisture contents, drying time, heat energy and electric fan energy and spoilage or storage index.

Optimization strategies Natural and heated air grain drying systems can be optimized by optimizing the total cost of

the drying systems within set bounds of drying time and spoilage. The manual optimization of the total cost of drying used in this study takes into account the operating cost (energy consumption) and fixed cost (annual costs of fan and motor, and the drying and storage structure). The optimal operation of existing dryers, which may be equipped with one type of heater or the other, is the focus in this study. Except for the cost of bin floor and ducts, the cost of the storage structure (bin) was, therefore, not taken into account in calculating the annual fixed cost for the drying system.

Broadly, the five optimization strategies investigated were:

(i) Fan and Fixed Air Temperature Rise Control: continuous fan operation with the gas or electric heater switched on continuously to raise the ambient air temperature Ta by a fixed amount of AT (5 or 10°C). The plenum air temperature Tp was given in this case by

Tp = T,d + AT. (14)

(ii) Fan and Heater Control: continuous fan operation with the fixed-power gas or electric heater switched on whenever the relative humidity of the plenum air RHp was greater than the single fixed set-point RHsr (50%, 70%), or whenever the plenum air temperature Tp was lower than the single fixed set-point Tsf (10, 20, 40°C).

(iii) Fan and Daytime Solar Heating: continuous fan operation with variable ambient air temperature rise by solar heating during the day.

(iv) Hybrid Heating: a modified form of the fan and daytime solar heating strategy was when the collector outlet air temperature was raised by a fixed amount AT before the air entered the dryer.

(v) Fan with no Heating: continuous fan operation at a fixed airflow rate with no extra heat added. This was the simplest strategy. The plenum temperature was the ambient temperature plus fan AT.

S C H O E N A U et al.: I N - B I N D R Y I N G O F C A N O L A G R A I N 49

For all the above control strategies, the set bound of drying time was 15 days, and the bound of spoilage was set at the spoilage index SI < 1.

Experimental procedure for verification of the drying model An experimental metal bin, 0.305 m (1 ft) diameter and 1 m high, was constructed and equipped

for drying and canola grain. The experimental data collected in drying remoistened canola grain in the bin were used for verification of the in-bin drying simulation model. The bin was equipped with fan, air heater and conditioner, controls and airflow, humidity and pressure measuring instruments. Temperatures at different locations in the bin, air duct and plenum were measured with type-T thermocouples. Grain and air temperatures were measured at the plenum floor and at 10 cm intervals in the bin. In order to remove grain from the bin to monitor the drying process and changes in the grain moisture content, small holes were drilled on the bin wall at 10 cm intervals and metal tubes were welded to these holes from which grains at different depths were removed at desired intervals. The external tubes had screw caps to prevent the grain or air from coming out from the bin when the grain was drying. The moisture contents for the grain were determined by the oven method according to the ASAE Standard $352.1 [17]. All the measurements were done precisely and, except for moisture content, the output of the various measurements was fed into a multi-channel data logger connected to a computer, which printed out the values of various parameters at 1 min intervals.

The canola cultivar Tobin was used in the drying experiments. Clean canola seeds at about 6% moisture content were purchased from a local seed supplier. The seeds were remoistened in a container to the desired initial moisture content for a test by adding predetermined quantities of pure water on a sufficient sample size to fill the bin. The plastic container was sealed and tumbled gently in a tumbling machine for at least 1 h after adding the water, to ensure uniform and complete mixing. The sample was kept at least for 24 h at room temperature before putting it into the bin for a drying test. The initial moisture content of the sample was measured at the beginning of each test.

For the small bin diameter, heat transfer in the radial direction could become significant, especially when the drying air temperature is elevated above ambient as in heated-air drying. The bin was, therefore, wrapped with 0.2 m thick fiberglass insulation.

R E S U L T S AND D I S C U S S I O N

Comparison of measured and simulated canola drying data Figure 1 shows typical measured and simulated moisture content data for remoistened canola

grain initially at 24% moisture content (wet basis) with the plenum air relative humidity and

e ~

25 - - ~--I't-n'h~l --I --1--1--1 --l--l--l--I --I ml --I --I ~1 --I~1

° ° ° ° ° ° ° ° ° ° ° ° o ' v

20 ÷ ' ; - ÷ , \ , \o \ \

\ \ \ , 15 - , \ N

l 0 -- * ' , | . . 'X+x~,, \ l \ t . . t Je.$.,. + .+ 0

I I I I I I I I I 1 3 6 9 12 15 18 21 24 27 30

T i m e (h)

* Simulated bottom layer m Measured bottom layer + Simulated 2nd layer • Measured 2nd layer I Simulated top layer Q Measured top layer

Fig, 1

50 SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN

temperature set at 30% and 30°C, respectively, and airflow rate at 0.18 m/s or 6.8 m3/min t. The simulated and measured data are for the bottom layer, second layer (100 mm from floor) and top layer. The maximum difference between measured and simulated moisture content under the test conditions was 1%, and the maximum difference between measured and predicted air and grain temperatures at the plenum, bottom layer and top layer was I°C. A comparison of the data indicated in Fig. 1 shows that the computer model adequately simulated the canola drying process.

Drying time and minimum fan energy and heat energy consumption for different energy systems

Drying time is an important factor in comparing the performance of various drying schemes. Timeliness could become a crucial factor in drying and marketing of canola. It might be necessary to dry the damp product as fast as possible in order to meet market demand. In addition, in many farming enterprises, the production capacity might be such that two or more full bin loads need to be dried in storage during one season. As a management policy to minimize capital investment, the drying equipment (fan, ducts, heaters, etc.) is rotated from one bin to the other until drying is completed for all bins. In this case, it would be necessary that drying is completed in one bin as fast as possible to allow enough time for others. A maximum drying time of 15 days seems to be reasonable in these situations as a drying interval of 15-30 days is considered normal in most in-bin drying operations, depending on the moisture content of the grain at harvest and loading.

Unheated air drying. Table l(a) lists the combinations of drying dates, initial moisture contents and minimum airflows and fan energy that would be required for complete drying of the product within 15 days without supplemental heat. The required minimum outflow and fan energy ranged from 1.5 to 2 mZ/min t and 9 to 26 M J/t, respectively, for all months and for all initial moisture contents (13, 16, 19%). In many cases, the drying was not completed even at 2,0 m3/min t airflow. When the airflow was increased to 3 m3/min t, the drying could still not be completed within 15 days. The simulation showed that there was no significant gain in drying time with increased airflow beyond 2 m3/min t as shown in Table 1 (b). On the other hand, the consumed fan energy increased drastically with increasing airflow beyond 2 m3/min t.

Heated air drying with natural gas, propane or electricity. Tables 2(a) and (b) list the simulated natural gas, propane or electric heated-air drying conditions that yielded minimum airflows and minimum heat energy consumption for complete drying to 10 and 8% "avg-dry" and "thru-dry", respectively, within 15 days. The corresponding electric fan energy consumed is also indicated in these tables. The optimum drying air conditions were keeping the plenum air relative humidity below 50% in August, raising ambient air temperature by 5 or 10°C in September and maintaining the plenum air temperature at 20°C in October for the "avg-dry" policy. Except for the case of 19% initial m.c. in October, the required minimum airflow in each drying condition was 1.0 m3/min t for the "avg-dry" policy. This airflow was almost half of that required for natural air drying for the same condition of complete drying within 15 days. For the "thru-dry" policy, the optimum drying air conditions were raising ambient air temperature by 5°C in August, 10°C in September and maintaining the plenum air temperature at 20°C in October. The optimum airflow was 1.0 m3/min t for all months at 13% initial moisture content, but the airflow had to be increased to 1.5 and 2.0 m3/min t for the 16 and 19% moisture levels, respectively. The consumed electric fan energy was about 2-6% of the corresponding heat energy for both drying policies.

Heated air drying by keeping the plenum air relative humidity below 70% and by maintaining plenum air temperature at 10°C produced comparable drying times as natural air drying in August and September. For the typical Prairie location, the mean daily temperature and relative humidity are 17.7°C, 60% and 11.3°C, 63,% in August and September, respectively. These drying conditions are, therefore, equivalent to unheated drying in these months.

Although maintaining the plenum air temperature at 40°C produced the shortest drying times in all cases, the heat energy consumption was relatively high (over two times higher in some cases than for raising the ambient temperature by 5°C). At 13 % initial moisture content, the specific heat energy consumption for 8% "avg-dry" and 1.0 m3/min t airflow was 97, 116 and 150 MJ/t and 113, 226 and 312 MJ/t in August, September and October, respectively, for the ambient temperature plus 5°C and plenum temperature maintained at 40°C drying conditions, respectively. The specific energy consumption and drying days for these two schemes and for other heated-air drying conditions at the same airflow and initial moisture content are indicated in Fig. 2. Although

SCHOENAU et al.: IN-BIN D R Y I N G OF CANOLA GRAIN 51

Table l(a). Combinat ions of drying dates, moisture contents and minimum airflow rates that resulted in complete drying within 15 days and without supplemental heat for "avg-dry" and " thru-dry- policies and SI < 1

Electric fan energy Drying days consumption (M J/t)

Initial Airflow Start m.c. rate "avg-dry . . . . thru-dry . . . . avg-dry . . . . thru-dry" date (%) (m3/min t) (10%) (8%) (10%) (8%) (10%) (8%) (10%) (8%)

1 August 13 1.5 7 14 12 - - 9 18 16 16 1.5 I1 15 15 - - 15 20 20 19 2.0 12 15 15 - - 21 26 26

1 September 13 2.0 9 15 13 - - 16 26 23 16 2.0 13 - - t - - - - 23 - - - - 19 2.0 15 . . . . 26 - - --

1 October 13 2.0 11 - - 15 - - 19 - - 26 16 2.0 15 - - - - - - 26 - - 19 2.0 . . . . 26 - - - -

tDry ing not completed within 15 days for blanked entries.

Table l(b). Drying times for various airflow rates with unheated air for 19% initial grain moisture content in

August

Electric fan energy Drying days (M J/t)

Airflow rate "avg-dry" (m3/min t) (10%) (8%)

"avg-dry" 0 0 % ) (8%)

0.5 20 33 9 13 1.0 15 19 14 15 1.5 13 17 17 20 2.0 12 15 21 26 2.5 11.5 14.7 25 32 3.0 11 14.5 27 38

maintaining the plenum air relative humidity at 50% drying condition had the least heat energy consumption, the drying days were highest and at 1.0 m3/min t airflow and 13% initial moisture content, except in August, the product could not dry to 8% "avg-dry" within 15 days under this condition. When drying time is a limiting management factor, Tables 2(a) and (b) provide a useful guide in selecting an appropriate heated-air drying condition that would result in minimum airflow and energy consumption for complete drying within the time limit specified.

Solar drying. Solar crop drying has been demonstrated to be cost effective [23, 25, 26] and could be an effective alternative to conventional artificial heated air drying systems, especially in locations with good sunshine during the harvest season. The mean daily global horizontal solar radiation is 19.4, 13.6 and 8.3 MJ/m2-day for August, September and October, respectively, for Saska- toon [27]. Many other locations in the Prairies receive comparable levels of insolation in these months. The moderate air temperature increases required for most crop drying situations can readily be attained with fiat-plate solar collectors. The air temperature increase through a collector varies with the collector area, airflow rate and insolation intensity. In practical applications, once the required airflow has been chosen, the collector area is determined from the average insolation data and required air temperature increase through the collector. In most crop drying situations, 1.0-1.7 m 2 of collector area per m3/min airflow would be adequate [24].

Figure 3 shows simulated typical hourly collector outlet air temperature rise above ambient and global horizontal insolation (TMY) at 1 m3/min t airflow in August, September and October in Saskatoon. During the day, the mean collector air temperature rise ranged from 16~C in August to 14°C in September and 12°C in October. The mean daily useful energy gains for drying by the collector were also 693, 550 and 434 M J/day in August, September and October, respectively, at this airflow rate. At 13% initial moisture content, the equivalent specific heat energy consumptions were 103, 89 and 75 MJ/t for 8% "avg-dry" at 1.0 m3/min t airflow.

52 SCHOENAU et al.: IN-BIN D R Y I N G OF CANOLA GRAIN

i

. . . , " t : t f"fl 8

t : h ~

. ~ ~ .~

~.~. ~

~ , ~

,.-, 0 ,

• ~ . ~ ~

o~

0 ~

"~.._

. . . ~

"~ t3

¢/~ '-o

. ~ , . O

...q

-t:~ . . °

t . . ,

~ o ~ o °

O O

~ o

t.., • ~ ~o

~ ~ |

• ~, o o

e q

.<

~ o

m N

SCHOENAU et al.: IN-BIN D R Y I N G OF CANOLA GRAIN 53

e. o

E

.-e

350 --

300 --

250 --

200 --

150

100

50

0

August

m September

October

Ta+5 Ta+|0 Tp >~20 i

Tp~40

Plenum condi t ion

RHp~< 50%

Fig. 2

The drying times for solar heating only were comparable to drying by continuously raising the ambient temperature by 5°C for the same initial moisture contents and airflows as indicated in Table 5. Adding 5°C to the collector output air temperature further reduced the drying times, but at the expense of drying with relatively very high air temperatures during daytime operation at peak insolation periods. The high air temperatures would lead to undesirable grain cracking. Under adverse weather conditions (low insolation and high night-time humidities), it would be necessary to have a gas furnace as an auxiliary heating source to the solar collector. Table 5 also indicates the drying days and heat energy consumption when a natural gas furnace is operated as an auxiliary energy source at night only by adding 5°C to the ambient air temperature. The application of this policy, as against continuously adding 5°C, would result in 50-60% reduction in heat energy consumption with a little increase in drying time.

Comparison of energy, drying and equipment costs for various energy systems

Apart from energy availability and operations convenience, the criterion for choice of an energy system for a drying operation depends on the drying energy and equipment costs. Table 4 shows a comparison of total annual heating and drying equipment cost, electric fan energy cost, heat energy cost and total drying and equipment cost for heated air drying by using natural gas, propane, electricity and solar energy, and ambient air drying in complete in-bin drying of canola grain within 15 days in September for 13% initial grain moisture content, 1 m3/min t airflow rate,

30 --

L.~ 25

20

!lo o

0

-- 900

.+ ----.+.. - - 800 -- . . ~ - - ~ ""'+.

..'* •-..-.-i. "" +",, -- 700

-- , ~ ~ ' ' t t . ( ".. -- 600 ¢'~

/ / / ,00 i ' [ i ' ' ' 300

-- ." " "" "~'~x" ~ ~ : 1 0 0 " " ". ",.-- 200

6 7 8 9 10 II 12 13 14 15 16 17 18 19 20

Local t ime (h)

* Temperature: August o Temperature: September "~'Temperature: October + Insolation: August • Insolation: September × Insolation: October

Fig. 3

54 SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN

,,.~ O0 I v

e ~

~o.ff _o

0 " .~

• ~ o . ~ . - ~ ..---,

• ~. :, .~ :~

• r " ~

~'E"

• - . ~ - : ~ ~ ~ . - .~ ~ • - . ~ ~ ~ ' ~

~ . ~ ,.~ Y : ~ _~.~ ~'~ .~' ~ ~.~' : ~ .~ ~ ~.~ _ ~ ..~'- o ~ ~ ~ ~ ~ '~ o

SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN 55

0 0

~_~ ~._= o~__

x a ~

~ '~ 8~

ca

P a ~

~.~

~.~

..-- -~

~_~.~ .~.~ ~

~ . ~ ~

~..=

el

e-

>-. ca ~

~ o ~

e-,

" 0

+++~==~o ~ ~ ' ~ +.~

~ ' ~

~ . ~

~:=_ - ~ 6 8

o

~o

. ~ o

~ ~oo-o

ga N ~

. ~ i ~' E-~ ~'~ ~ ,

~ 0 0 0 0

~ - ~

o o 0

- ~

0 0 ~

0

oo c~

: ~ ~=

~ . . . . . . .

• + . ~ + II II 11

~ ~

. o

< • 0

,.1=

g

E

E

o

56 SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN

and 10 and 8% "avg-dry" policy. The drying air condition for the natural gas, propane and electric energy systems was ambient temperature +5°C, collector outlet air temperature and collector outlet temperature + 5°C for the solar energy and hybrid systems and ambient temperature only for the natural or unheated air system. For the unheated air system, the airflow rate had to be increased to 2 m3/min t to complete drying within 15 days.

The cost comparison for heated-air drying in Table 4 showed evidently it was cheapest to use solar energy only with a total drying and equipment cost of $5.36/t and $5.42/t, followed by the natural gas system with $6.54/t and $6.75/t, and propane system with $6.56/t and $6.78/t for 10 and 8% "avg-dry" policy, respectively. The least cost effective was combined solar energy and natural gas system followed by the electric air heater system. Although drying with unheated air had the least total drying and equipment cost, the airflow had to be increased by two times, with a high possibility of drying uncompleted within 15 days and mold attack on the top grain layers. The heat energy cost for the natural gas system was only $0.02/t to $0.03/t cheaper than the propane system, while the electric heat energy cost was about 4 times higher than natural gas or propane energy. When the choice in a drying operation is for conventional energy sources, it is, therefore, best to use natural gas or propane. In locations with favorable solar energy availability, it is also cost effective to use a solar energy air heating system with a little trade-off on handling convenience and reliability.

Optimum drying air conditions and minimum drying costs

Average-dry policy. Table 5 lists the optimum drying conditions that yielded minimum drying cost (including equipment cost) for 10 and 8% "avg-dry" policy for complete drying in 15 days or less. Unheated air could be used for complete drying in August with airflows ranging from 1.5 to 2 m3/min t, but supplemental heating with ambient air temperature raised by 5 or 10°C and plenum temperature maintained at 20°C resulted in complete drying and minimum costs in September and October, respectively. For all months, the minimum drying costs increased with increasing moisture content.

Solar heated-air drying was also comparatively cost effective in conditions where unheated air drying was not adequate in September and October. Generally, the cooler outdoor temperatures in September and October slowed down the deterioration of the undried portion of the stored grain in the bin and allowed complete grain drying with supplemental heat at lower airflows. The minimum drying costs ranged from $4.31/t to $4.98/t for drying with unheated air in August, $6.54/t to $7.23/t, and $6.41/t to $8.06/t in September and October, respectively. Increasing the airflow rate slightly increased drying cost. For example, in August at 16% initial moisture content, by increasing the airflow from 0.5 to 2 m3/min/t in unheated air drying, the drying cost increased from $3.26/t to $4.96/t, while the savings on drying time was more than 2 times at the higher airflow rate.

Through-dry policy. Table 6 lists the optimum drying conditions that yielded minimum drying costs for 10 and 8% "thru-dry" policy and complete drying within 15 days in August, September and October for 13, 16 and 19% initial moisture contents. The corresponding over-drying costs, and total drying and over-drying costs are indicated in the table. Unheated air drying was the optimum condition for drying in August at initial grain moisture content equal to or below 16% and relatively high airflow of 2 m3/min t. The drying cost was between $4.28/t and $4.46/t, while the corresponding over-drying cost ranged from $1.44/t to $2.96/t.

To dry 19% m.c. grain in August, the ambient temperature had to be raised by 5°C. For drying in September and October, the optimum drying conditions were raising the ambient temperature by 10°C and solar heating only, and maintaining the plenum temperature at 20°C, respectively. For these heated-air drying conditions, the drying cost ranged from $6.14/t to $8.57/t, while the corresponding over-drying cost was between $1.72/t and $6.02/t. The least values for these costs were obtained for the solar heating system. The over-drying costs generally were lower for the unheated air drying condition than for the heated-air schemes. The over-drying cost also decreased with increasing airflow rate.

For the "thru-dry" policy, the over-drying cost was higher in drying to 10% than to 8% final moisture content. This is because the calculation of over-drying weight loss was based on the

S C H O E N A U e t al.: IN-BIN D R Y I N G OF C A N O L A G R A I N 57

" V , =

~ 6

>,.6

0 2 :

02

8 -

• ~, ~, v

• ~ . 6 : ~~ .~ 02 ~

. ~

N ~

~8

~ 02

._~ N ~ 02

02 ~ .

'N ~ .

N ~

[-.. ~

• B'oq " o " ~

.6 . ~

02 O

e ~ q

e ~ O 0 2 ~

O

O

,-d ,d ,-,q

t "~ ~D

¢ ~ ' q D ,,D tt-~ r-,- o ~

r-.- o o

o o r - - ¢,q ¢,~

o O

¢-,I e ~

~ O

~t~ 02

¢/1 , . ~

: : 6

e~ 02 e~

m ~ m

58 SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN

difference between the actual final average moisture content after drying and the desired final average moisture content (10 or 8%), not on an arbitrary value. The actual final average moisture contents were closer to 8% than to 10%. Over-drying, no doubt, constitutes a major concern in through-drying. As the slow movement of the drying front and over-heating are the major causes of over-drying in deep-bed drying, a possible optimum control strategy to reduce the over-drying cost would be to ensure rapid drying initially by adding extra heat in the periods of high ambient relative humidity and later when drying is advanced, the extra heat is reduced gradually. This control strategy would obviously reduce over-heating, over-drying, and total drying and over-dry- ing cost. In through-drying, control strategies that would lead to a reduction in over-drying are more important than controls for optimizing the heat energy consumption.

CONCLUSIONS

(1) Continuous fan operation with 1.5-2 m3/min t ambient air is required to dry canola grains to l0 and 8% average-dry and through-dry moisture contents in 15 days or less in August at 19% initial moisture content or less. Supplemental heat must be applied to successfully dry the product in September and October.

(2) In humid weather conditions, where the air relative humidity is continuously higher than 70%, supplemental heat is required to raise the plenum air temperature a few degrees (typically 5-10°C) and lower relative humidity as high air humidities lead to rapid grain spoilage.

(3) Since unheated air drying conditions, requiring airflows higher than 1.5 m3/min t and installation of large capacity fans, may not be feasible in many farm locations with limited electricity supply, drying with supplemental heat is more appropriate in these conditions, as it may be less costly to dry with supplemental heat at 1 m3/min t airflow than drying with unheated air at the increased airflow of 2 m3/min t.

(4) The drying time is about 60-100% or 20-50% higher for drying to l0 or 8% moisture contents, respectively, when the drying front is allowed to sweep the entire grain bed (through-dry) as compared to drying grain to a bin average moisture content (average-dry).

(5) The average-dry policy is generally more economical than the through-dry policy. Although the labor cost for mixing grains after drying makes the drying cost for the average-dry policy about 3-10% higher than for through-dry, the combined drying and over-drying costs for through-dry resulted in 30-100% higher costs. However, for grains that must remain in storage for a few months, generally more than 3 months after drying, the through-dry policy is more appropriate as all layers of the grain are dried to a safe moisture content in this way.

(6) Continuous fan operation in August and September with supplemental heat by natural gas or propane to raise the ambient temperature by 5-10°C and solar heated-air drying resulted in the least drying costs among the supplemental heat systems. In October, solar heating and maintaining the plenum air temperature at 20°C with a natural gas heater resulted in the least drying costs.

(7) Solar drying has been found to be more cost effective than other supplemental heat systems provided a well designed flat-plate solar collector for air heating can be found for use in locations with good solar energy availability. With a proper design, flat-plate solar heaters may be readily incorporated in existing or new in-bin drying systems. A solar collector area of 1-1.2 m 2 per m3/min of airflow is adequate for most drying situations.

(8) Heating the drying air with natural gas or propane is the most cost effective when the use of conventional energy sources is preferred to or more convenient than renewable energy sources in grain drying operations.

R E F E R E N C E S

I. S. Sokhansanj, E. A. Arinze and G. J. Schoenau, Management schemes for optimum control of inobin drying of canola. Project Report No. 9037, College of Engineering, Univ. of Saskatchewan, Saskatoon, Canada (1992).

2. L. A. Appelquist and B. Loof, Postharvest handling and storage of rapeseed. In Rapeseed (Edited by L. A. Appelquist and R. Ohlson), pp. 60-100. Elsevier, Amsterdam (1972).

SCHOENAU et al.: IN-BIN DRYING OF CANOLA GRAIN 59

3. E. A. Arinze, G. J. Schoenau and R. W. Besant, Sol. Energy 36, 191 (1985). 4. T. L. Thompson, R. H. Peart and G. H. Foster, Trans. ASAE 11, 582 (1968). 5. S. Sokhansanj, W. G. Lang and D. E. Lischynski, Can. Agric. Engng. 33, 265 (1991). 6. ASHRAE, ASHRAE Fundamentals Handbook. ASHRAE, Atlanta, Ga (1990). 7. E. A. Arinze, S. S. Adefila and S. N. Mumah, Niger. J. Sol. Energy 8, 103 (1989). 8. J. A. Duffle and W. A. Beckman, Solar Engineering of Thermal Processes. Wiley, New York (1980). 9. ACAE, Barn Hay Drying. ACAE Publication No. 8, Atlantic Committee on Agricultural Engineering, Atlantic

Provinces Agricultural Services Coordinating Committee, Halifax, Canada (1986). 10. P. Shatadal, D. S. Jayas and N. D. G. White, Trans. ASAE 33, 871 (1990). 11. W. E. Muir and R. N. Sinha, Can. Agric. Engng. 28, 45 (1986). 12. S. Sokhansanj, W. Zhijie, D. Jayas and T. Kameoka, ASAE 29, 837 (1986). 13. C. C. Chen and R. V. Morey, Trans. ASAE 32, 983 (1989). 14. M. E. Nellist and C. M. Bruce, Drying and storage of oilseed rape in the UK. Part I. Physical and engineering aspects.

Report No. CR/453/92/8860, Silsoe Research Institute, Agricultural and Food Research Council, Silsoe, U.K. (1992). 15. E. A. Arinze, S. Sokhansanj and G. J. Schoenau, Development of optimal management schemes for in-bin drying of

canola grain (rapeseed). Computers Electron. Agric.: An. Int. J. In press. 16. W. E. Muir, R. N. Sinha, Z. Zhang and D. Tuma, Near-ambient drying of canola. Paper No. 89-6532, ASAE, St.

Joseph, Mich. (1989). 17. ASAE, ASAE Standards. Am. Soc. Agric. Engrs. St. Joseph, Mich. (1990). 18. D. S. Jayas and S. Sokhansanj, Trans. ASAE 32, St. Joseph, Mich. (1989). 19. E. A. Arinze, G. J. Schoenau, S. Sokhansanj, S. S. Adefila and S. M. Mumah, Energy Convers. Mgmt 34, 33 (1992). 20. A. J. de Ron, Sol. Energy 24, 117 (1980). 21. J. Kreyger, Drying and storing grains, seeds and pulses in temperate climates. Instituut Voor Bewaring, En Verwverking

Van Landbouwprodukten, Pub. 205 (1972). 22. E. Audsley and D. S. Boyce, J. Agric. Engng Res. 19, 173 (1974). 23. Brace Research Institute, A survey of solar agricultural dryers. Tech. Rept T-99, Brace Research Institute, Macdonald

College, Quebec, Canada (1975). 24. B. Brenndorfer, L. Kennedy, C. O. O. Bateman, D. S. Trim, G. C. Mrema and C. Wereko-Brobby, Solar dryers--their

role in post-harvest technology. Commonwealth Secretariat, London, U.K. (1987). 25. R. O. Pierce and T. L. Thompson, Solar grain drying in the North Central Region--simulation results. Paper No.

76-3517, ASAE, St. Joseph, Mich. (1976). 26. E. M. Wrubleski and P. S. Catania, Experience with solar collector for grain drying in Saskatchewan. Paper No. 77-106,

Can. Soc. Agric. Engrs, Guelph, Ontario (1977). 27. J. L. Bergsteinsson and J. G. Calvert, Saskatoon climatological reference station annual summary. Saskatchewan

Research Council, Saskatoon (1985).