Development of Time-Temperature Probes for Tracking Pathogen ...
http://www.redorbit.com/news/science/1492793/devel [2008-7-25]
Tag : Metal Locator
In order to account for the effects of initial location within thepile on probe movement, the probes were placed in three differentlocations, roughly at the top, middle, and bottom of a windrow atthe co-composter cure site. One third of the probes of each densitybeing tested were placed at each height. All probes were placed inthe same lateral position in order to minimize the search areaafter turning. The heights of the probes in each position weremeasured and recorded. The windrow was then turned once and theprobes recovered. This was accomplished by digging through thecompost by hand, in order not to affect the location of the probesduring their recovery. When a probe case was located, its finalheight was recorded. The second run was carried out approximately 4weeks after the first run, in the same windrow as the first trial.The procedure was similar, with the exception of the number ofcases introduced, the method of securing the end caps to thealuminum tube, and the densities tested. For this trial, 27 casesof each of three densities were introduced into the windrow. Theend caps were secured to the aluminum tubing using electrical tapeonly, as duct tape tended to tear during the first trial. The highdensity was changed to 2000 kg/m^sup 3^, which is slightly higherthan the density of the commercial temperature logger. The lowdensity was kept at 800 kg/m^sup 3^, and thus the midpoint was 1400kg/m^sup 3^. The remainder of the procedure was the same as duringthe first trial, with one third of the probes of each density beingplaced at the top, middle, and bottom of the windrow, the windrowbeing turned, and the location of the probes after turning beingrecorded.
During the second run, the wet bulk density of the compost wastested, in order to have a point of comparison between probe andcompost densities. Compost was collected from four locations (top,bottom, and two middle locations) in the windrow used for testing(after turning), in the same zone as the temperature probe caseswere placed. These four samples were amalgamated in a 5-gallonbucket with a lid and stored for 10 days prior to analysis. Theprocedure used for the bulk density determination was amodification of TMECC Method 03.03-A/03.01-A (steps 10.3 to 10.5only) (TMECC 2001).
Change in height data (i.e. the difference between initial andfinal heights) was analyzed using a two-factor analysis of variance(ANOVA), with density and initial height being the factorsconsidered. The effect of density on probe movement was of majorconcern, but the initial height of the probes also needed to beconsidered because the starting position affected the possibledirections (up or down) and magnitudes of the height change. TheANOVA analysis provided insight into whether probe densitysignificantly affected the change in height of the probes afterturning.
The cross-sectional dimensions of the windrow were measured (i.e.width at the top, w^sub 1^, width at the bottom, w^sub 2^, andheight, h) (see Figure 1). From these dimensions, it was possibleto approximate the distance, d, above ground at which an equalvolume of compost lay above and below. Since probes were initiallyplaced in equal numbers at the top, middle, and bottom of the pile,the corresponding change in height, Deltah, to this volume midpointshould be the difference between this distance, d, and the half thepile height (i.e. the mean initial location). It was postulatedthat random probe movement during windrow turning would beindicated by a mean change in height corresponding to the volumemidpoint, since there would be a 50% chance that material would layabove or below this line. A t-test was used to evaluate this.
FIGURE 1. Approximation of a windrow cross-section; dimension "d"indicates the level in the pile where the volume of material aboveand below the line is approximately equal
Robustness of Circuitry
It was also of prime importance that the temperature loggingcircuit inside the housing remain intact and functioning for theduration of the data-logging mission, regardless of the conditionsencountered. Thus, during tests of density and housing materialstrength in the co-composter cure site windrow, some quickexperiments were also done to examine the robustness of theelectronics to severe impacts (as occur during windrow turning).
During the first run, an operating commercial temperature loggerwas introduced into the middle of the windrow, as was anoperational DS2422-based temperature logging circuit with a singleAA battery. The DS2422 circuit was enclosed in a section ofaluminum tubing in such a way that minimal movement of the circuitboard would occur inside the housing, and the tube was capped witha 50.8 mm diameter, 1.58 mm thick polycarbonate circle on each end.The windrow was turned, the circuits recovered, and the electronicsexamined for damage.
During the second windrow-turning trial, in addition to the twotemperature logging circuits listed above, six cases containing 1/2AA sized lithium batteries soldered onto a circuit board wereplaced into the windrow at the bottom. Three of the cases containedtabbed batteries (TL-5902/T) and the other three containedbatteries with solder pins (TL-5902/P). Again, the circuits wererecovered after windrow turning and examined for damage.
Temperature Response Time
The response time of a temperature sensor is dependent in a numberof factors, including the thermal conductivity of the sensor itselfand of the materials (such as the case walls and air space) betweenthe sensor and the temperature of interest. It was desired that theamount of time required for both the commercial and DS2422- basedloggers to respond to a step change in temperature be determined.An excessively long response time would be undesirable becausetemperature fluctuations of interest may be overlooked; three hoursor less was set as the target.
A two-point temperature calibration (at approximate temperatures of-10[degrees]C and 60[degrees]C) was done for each device, to ensurethat both would provide the same data under equivalent conditions.The two-point calibration was completed as specified in thedocumentation for each device (Madge Tech n.d.; Maxim/Dallas 2003).After calibration, each device was set up to record temperatureonce every 60 s and enclosed in its housing (the DS2422 circuithousing was as described in section 4.4.3, except that the end capswere 4.76 mm thick). The temperature response times of both deviceswere evaluated over temperature changes of approximately +-10[degrees]C, +-15[degrees]C, +-20[degrees]C, +-30[degrees]C, and+- 40[degrees]C. These changes were achieved by moving the probesbetween a room temperature zone and a pre-heated lab oven. When thedevices were moved from one temperature zone to another, they wereallowed to equilibrate for a period of at least 2 hours (based onpreliminary observations), to ensure that sufficient time had beenallowed for the temperature readings to stabilize. The timerequired for a given temperature rise or fall was evaluated whenthe device had attained a temperature within 0.5[degrees]C of itspeak or minimum temperature, respectively.
After all specified positive and negative temperature change cycleshad been completed, temperature data was downloaded from eachdevice and evaluated to determine the response time for each stepchange in temperature.
This test was performed at the University of Alberta.
Probe Recovery
It was desired that as close as possible to 100% of the probesintroduced into a composting system be recovered, so as not to losecollected data. Several probe detection and recovery methods wereconsidered during the design phases. Of those considered, visualrecovery during screening and location by metal detection wereconsidered the most viable options.
The efficiency of probe recovery during screening was tested byintroducing 30 sections (six each of the following colors: gold,blue, yellow, orange, and unpainted) of aluminum tubing into arotary drum screener along with a loader bucket of compost. Theovers of the screen were scanned visually for the aluminum tubesand all those spotted were collected and counted. The differentcolour tubes were evaluated for relative visibility levels. Thistest was repeated twice: once in the finished product of theGore(R) composter and once in compost from the co-composter.
Two trials were also done to determine how easily the temperatureprobes could be located using metal detection. First, the ease oflocating the aluminum sections with a relatively basic, low-costhand-held metal detector (Garret Treasure Ace 100) was tested. Themaximum depth that the aluminum probes could be sensed with themetal detector was determined by burying a piece of aluminum tubingin screened biosolids compost at a known location at a known depth,varying from 1 cm to 20 cm. The metal detector was set on its mostsensitive setting and the compost was scanned. The depth at whichthe piece of aluminum could no longer be sensed was noted. It wasalso of interest to see how easy it would be to locate a stainlesssteel probe (in case the housing material was modified) and to seehow the performance of an industrial detector would compare withthe hobby model used in the first test. Thus, a magnetic "pinfinder" (Schonstedt GA-52Cx Magnetic Locator) was used to locatemagnetic stainless steel rods buried in a co-composter cure pileand also in a yard waste pile. A similar test to that done with theGarrett Treasure Ace was performed.
Effect of Magnetic Field on Data Memory
In some composting operations, magnetic separation is employedeither at the front-end or during screening of the final product.Such operations use strong magnetic fields to separate out andcapture ferrous materials. It is therefore possible that thetemperature loggers may enter into a relatively strong magneticfield during their use. It was of interest to determine whateffect, if any, a strong magnetic field would have on a device, asdata loss or any change in functioning would be undesirable. Theeffect of magnetic fields is not a standard test for mostsemiconductor devices (TI Support 2005), and was therefore unknown.
An experiment was conducted on both the commercial device and theDS2422-based circuit. A temperature-logging mission was started oneach device, and data was collected for one day. This data wasdownloaded and saved without stopping the data-logging missions orerasing data. The devices were then enclosed in their cases and putunder the magnetic separation unit at the EWMCE's co-composter fortwo minutes (which exceeds the amount of time it would normallytake for an object to pass under the magnet). Data was downloadedfrom each device and compared to the data downloaded prior tomagnet exposure. A second temperature-logging mission was startedon both devices and the data from that mission examined to ensurethat device functioning was not affected. Results
Material Strength and Robustness
None of the 16, 14, or 11-gauge aluminum tubes suffered any damage,aside from minor paint chips, when they were dropped from a loaderbucket onto a packed gravel surface. When driven over by afront-end loader the 14- and 16-ga. tubes were crushed, while the11- ga. tube was compressed only slightly (Figure 2). It isimportant to note that this experiment was conducted without therecommended end caps in place, which may have made the casesstronger by providing structural support at the ends.
During the windrow-turning tests, aluminum sections of all gaugeswere dented by the turner's auger blades. Some of the most severedamage can be seen in Figure 3. While all gauges of aluminum wereaffected to some degree, the most acute damage occurred to the16-ga. tubing and the least amount of severe denting occurred to the11- ga. tubes. More and larger dents tended to occur at the ends ofthe tubes rather than in the middle. Again, it should be noted thathaving the recommended end caps in place may have provided someadditional structural support and hence reduced the severity of thedenting at the tube ends.
FIGURE 2. Damage to aluminum tubes driven over by a front-endloader. a) 16-ga. tube. b) 14-ga. tube. c) 11-ga. tube.
FIGURE 3. Examples of damage to aluminum tubes after windrowturning, including denting and deformation. a) 16 ga. tubes. b) 14-ga. tubes. c) 11-ga. tubes.
FIGURE 4. Chemical and microbiological degradation of aluminumhousings recovered after four weeks in a compost cure pile (darkspots indicate areas of damage): a) unpainted, unanodized tube andb) anodized commercial logger case.
Qualitative observations of damage sustained to the aluminum casesdue to chemical and biological factors were also made during thewindrow test. Some of the painted cases darkened in colour afteronly a few days in the cure pile. The gold cases were beginning toshow signs of corrosion or microbial degradation of the paint andaluminum after having been in compost for less than four days. Oneunpainted case was recovered during the second trial after spendingnearly a month in the windrow; this case was severely pitted innumerous locations due to some combination of corrosion andmicrobial degradation. The housing of the commercial temperaturelogger, made from the same aluminum grade, but anodized, was notaffected by chemical or microbiological stresses after the sameamount of time in the cure pile (see Figure 4).
Probe Density
During the first trial, a significant amount of data was losteither because the probe housing was not recovered or because oneor both ends of the case fell off, resulting in a significantdensity change. Out of 26 housings introduced to the windrow, only20 were recovered. Of these, 9 had been broken during turning andsuffered a change in density, leaving only 11 probes intact. Thus,only 11 probes provided useful information for data analysis. Atwo-factor ANOVA was performed on these 11 data points for changein height data. The results of the analysis indicated that, oncethe initial location of the probes was taken into account, therewas not a significant difference in height changes due to differentdensities. However, since there was only a small amount of usefuldata obtained during the first trial, no strong conclusions couldbe drawn regarding the affect of density on probe movement.
A second trial was performed because so many probes were lost ordamaged during the first trial. Three times as many probe housingswere used and the end caps better secured. During the secondexperiment 7 probes were not recovered and two broke. However, thelarge amount of remaining data meant that a more meaningfulanalysis could be completed than was possible after the firsttrial. The dimensions of a cross-section of the windrow weremeasured (as in Figure 1). The pile height, h, was around 1.5 m.The width of the windrow at the top, w^sub 1^, was approximately1.2 m. The width at the bottom of the pile, w^sub 2^, was roughly5.5 m. From these dimensions, it was determined that there wereapproximately equal volumes of material above and below d = 0.53 m,which corresponded to a change in height, Deltah, of -0.22m.Compost bulk density during this trial was 446.9 +- 8.7 kg/m^sup3^.
From the average change in height data for each density, itappeared that there was a slight trend toward more downward probemovement as density increased. Interestingly, a best-fit line(linear regression) through the average change in height datacrossed the volume midpoint (Deltah = -0.22 m) at a density of 452kg/m^sup 3^, which is nearly the same as the compost bulk density.However, when the data was weighted to take into account thedifferent numbers of probes that were lost from each height, thetrend toward downward movement with increasing density was not asobvious (Figure 5) and the volume midpoint was crossed at a densityof 246 kg/m^sup 3^. Additionally, after accounting for the effectsof initial probe location, an ANOVA analysis (95% confidence level)indicated no significant difference in change in height databetween probes of different densities. In other words, the ANOVAresults indicate that varying probe density from 800 kg/m^sup 3^ to2000 kg/ m^sup 3^ appeared not to significantly affect probemovement.
FIGURE 5. Plot of change in height data after windrow turning (raw,averaged, and weighted average) for probes of different densities.
T-tests indicated that for all densities, neither the average northe weighted average change in height significantly differed fromthe compost volume midpoint. This result, along with the ANOVAresult that there was no significant difference in change of heightbetween different densities, seems to indicate that there wasrandom dispersion for probes with densities between 800 and 2000kg/m^sup 3^ during turning of a windrow with a bulk density near450 kg/ m^sup 3^.
Robustness of Circuitry
During the first trial, only the DS2422 device was recovered afterwindrow turning. All circuit components and solder contactsremained intact during windrow turning. However, one of the tabstore off of the battery, so that power to the circuit was lost (seeFigure 6). When the battery was replaced the circuit functionedproperly; however, no data was retained. During the second trial,several 1/2AA-sized batteries of both tabbed and solder pinvarieties were tested in order to see if the battery problemsencountered during initial testing would occur again, and to see ifpins would be stronger than tabs. After turning, the commerciallogger was recovered along with the DS2422 circuit, all three casescontaining tabbed 1/2AA batteries, and one of the three casescontaining batteries with solder pins. All of the solderedbatteries had one or both tabs/pins broken. Again, the onlyapparent problem with the DS2422 circuit was the broken battery,and the device functioned normally when it was replaced.
FIGURE 6. Damage to the DS2422-based temperature logging circuitafter being turned once by a windrow turner. One of the batterytabs was torn off.
The battery in the commercial temperature logging device wassimilar to the other 1/2AA batteries with pins that were tested,except that it was placed in a socket rather than being soldereddirectly into the circuit. Sometime during the three windrowturning activities experienced by the commercial logger (one duringthe initial trial, one during the second trial, and one betweentrials), one of the battery pins came out of its socket and thedevice lost power. Another main component of the circuit, amicrocontroller, also popped out of its socket, and a diode wasbroken in half (see Figure 7). It was hypothesized that the diodebroke because it was hit when the battery came out of its socket.All other components of the commercially available device weresoldered onto the circuit board and survived three passes of thewindrow turner. When all broken or loose parts were replaced, thelog-ger resumed normal operation. Data memory was full, indicatingthat damage to the device likely occurred during the third turningevent.
Figure 7. Damage to the internal circuitry of the commercialtemperature logger after being turned three times with a windrowturner. One of the components came completely out of its socket.
TABLE 2.
Time required for each of the commercial and DS2422-based devicesto respond to different changes in temperature.
Temperature Response Time
The results of the temperature response test are presented in Table2. The commercial and DS2422-based temperature loggers had similarresponse times. The maximum amount of time needed for either deviceto respond to a step change in temperature was 81 minutes, whicheasily met the desired three-hour limit set on response time.Though the data shows an overall trend of longer response times forlarger changes in temperature, the temperature probes are notexpected to experience a step change in temperature much more thanthe largest change tested. Therefore, the temperature responsetimes of both devices were deemed adequate.
Probe Recovery
When various colours of aluminum tubing were screened with bothbiosolids and mixed biosolids/ MSW compost, a person visuallyscanning the screen overs easily recovered 100% of the tubing. Inthe biosolids compost, all colours were equally easy to see.However, in the mixed compost the gold and unpainted cases tendedto blend slightly more than the other colours with non-compostablematerials (such as bits of coloured plastic, glass, and metal) inthe product. Nonetheless, it appeared that screening the probes outof finished compost would be a viable method of recovery. The low-cost metal detector used to evaluate the recovery of aluminum probecases performed quite poorly. A case placed at a known locationcould only be detected up to a depth of 15 cm. Numerous otherobjects were also picked up by the metal detector (even in screenedbiosolids compost). In fact, other objects were picked up with suchfrequency that an operator would be unable to determine whether thedetector was signalling the presence of a probe or of someextraneous object. Reducing the sensitivity of the detector was noteffective in reducing this interference. Additionally, with reducedsensitivity the maximum detection depth decreases.
The magnetic locator was also found to be ineffective for use inprobe recovery. In the co-composter cure pile, a stainless steelbar was easily detected, but other objects in the pile were alsopicked up. As was the case with the metal detector used to locatealuminum tubing, it would be virtually impossible to distinguishbetween a stainless steel probe and other ferrous objects in thecompost. The situation was even worse in the yard waste windrow,with the detector giving strong signals whether or not thestainless steel bars were in the pile.
Effect of Magnetic Field on Data Memory
After exposure to a magnetic field generated by a ferrous magneticseparation unit, temperature data collected by the DS2422 deviceand the commercial device was compared to that collected prior tothe exposure, and was found to be the same. Data collectioncontinued, unaffected, during and after subjection to a magneticfield. The devices also both operated normally when they werestopped and started again, indicating that a strong magnetic fieldhad no effect on the operation of either device.
Discussion and Recommendations
Testing of the designed temperature logging device and a similarcommercially available model revealed that, while both devicesshowed potential as far as performing their desired function,neither device was ideal. Improvements could be made to bothdevices in order to obtain a better fit between the specifiedparameters and the actual performance of the probes.
One of the most important features of the temperature probe was anability for the devices to mimic the behaviour of random particlesof compost during agitation and settling. The average position of aprobe should tend toward the volume midpoint of a compost pile(i.e. there should be a random distribution of probes around thelevel at which there are equal volumes of compost in the top andbottom of the pile). Three probe densities were tested (800,1400,and 2000 kg/m^sup 3^) in compost with a bulk density ofapproximately 450 kg/m^sup 3^. An ANOVA analysis indicated nosignificant differences between the different probe densities, andthere were no statistically significant difference between theaverage change in height data and the volume midpoint for all probedensities. These results imply that random probe distribution wasachieved. However, they were based on a single experiment;additional trials are recommended to confirm the results. It shouldalso be noted that even though random distribution was observedduring this trial, the same results may not be seen with probes ofdifferent sizes or in a pile with fresh feedstock materials. Thisis because during this test the probe size was significantly largerthan the average pore size within the compost heap and thus theprobes were easily held up by the compost matrix. If the probe sizeapproaches the size of a compost pore, it will likely become easierfor dense probes to move downward in the pile. Hence, if the probedimensions are decreased such that they become similar to poresize, this test should be repeated.
Edge effects should also be considered in future trials todetermine how well the probes are able to mimic random probebehaviour. Such effects may include an inability for a windrowturner to pick up a small probe sitting right at the ground (thoughin this case the probe would be able to monitor conditions seen bya compost particle sitting at ground level). Also, since the probesare cylindrical, if they reach the edge of the pile there is a highprobability that they will roll off the pile, whereas a compostparticle may not. This is an undesirable situation, as valuabledata could be lost. A possible way to prevent rolling would be toattach a counterweight or hook to the device.
It was also important to be able to recover all (or nearly all) ofthe probes introduced into the compost. Screening appeared to be afeasible method of probe recovery. However, it is recommended thatadditional screening tests be performed after probes have undergonea complete compost cycle, as it is possible that the compression ofmaterials within a pile could cause compost to stick to the casesand thus make them difficult to see. The results reported in thispaper refer to a screening test done immediately after adding casesto relatively dry compost. Two attempts at probe recovery by metaldetection failed, partly because detection depth was small andpartly because there was interference from extraneous objects inthe compost pile. It is possible that a more expensive detector mayperform better; some higher end detectors have visual displayswhich allow specific objects to be identified by signal/wave shape.These detectors are also able to penetrate deeper, though maximumdepths of only a few feet are reported, which may still be tooshallow (Catalano 2006). At this point, screening seems to be themost viable probe recovery method. Since neither device appeared tobe affected at all by exposure to a strong magnetic field, it isnot of concern if screening operations employ magnets to removeferrous materials from the compost.
The temperature response times of both the commercial and DS2422-based temperature loggers were satisfactory; the response timeswere short enough that it is unlikely any significant variation intemperature (with time) would be overlooked. However, if majorchanges are made to the probe design, these tests should berepeated to confirm that the results are still adequate. Changes tothe amount of air space inside the probe or changes to thematerials used could potentially affect the amount of time requiredfor the device to respond to a temperature change.
As far as the probe housing is concerned, it was seen that 16-gauge aluminum was not sturdy enough to survive relativelyunscathed during activities such as windrow turning. It isrecommended that a thicker gauge, such as 14 or 11, be used in thefinal design. The 11- gauge tubing did suffer some denting duringturning, but the severity of the dents was small relative to thoseon the other two gauges. However, since 11-gauge adds extra mass tothe probes 14- gauge might be a better choice since it survivedreasonably well during high impact situations and would contributeless to increasing probe density. It was seen that dent frequencyand severity was greater at the ends of all gauges of aluminumtubes than in the middle. Severe denting at the tube ends is causefor concern, given the proposed design of the case (an aluminumtube with polypropylene end caps fitting inside the tube); it ispossible that even minor dents at the tube ends tubes may damagethe seal between the end caps and the aluminum tube. This couldlead to exposure of the electronic circuitry to moisture, compost,and/or contaminants. It is particularly important to preventmoisture from coming into contact with the lithium battery used, asthis may present an explosion hazard. It should be noted, however,that had the suggested end caps been in place during testing, theremay have been additional structural support and dent severity mayhave been reduced.
A brightly coloured case (i.e. blue, yellow, or orange) isrecommended to increase visibility. Aluminum can be anodizeddifferent colours; this is advised, as anodizing seemed to beeffective in preventing corrosion and would limit paint chippingand discoloration.
The most serious problem encountered during windrow turningexperiments was that electronic components and batteries placed insockets popped out of the sockets, and batteries soldered to thecircuit board suffered broken pins or tabs. The ability of thedevice to function as desired was affected when power was lost and/or circuit components failed. The use of solder-mount chips ratherthan sockets would solve half the problem. However, anothersolution is needed for the problem of battery breakage.
It is hypothesized that battery breakage occurred due toconservation of momentum; when the probe case was accelerated ordecelerated, the battery on the inside of the case did notexperience the same change in momentum. In other words, when thecase, and hence the circuit board, was stopped, the battery wasable to keep moving and the solder tabs broke. This was a problemfor the battery but not for other soldered parts because thebattery was relatively large and heavy (momentum is a function ofmass) and because the battery was attached to the circuit board inonly two places (at either end), while other parts were attached atseveral points along two edges. It is important to find a solutionto the problem of battery breakage, as this causes power loss andhence data loss. The solution is probably as simple as finding abetter way to secure the battery to the circuit board. One optionwould be to create a mould that would fit snugly around the circuitboard and battery, preventing movement of the internal parts.Another option, suggested by the manufacturer of the commercialdata logger, would be to wrap tape around the battery to keep it inplace; electrical tape may be suitable for this purpose, as it wasstrong enough to hold end caps in place during windrow turning. Thetape option may be preferable because the temperature response timemay be increased by the mould material. Tape would also add lessweight to the probe. Testing is recommended prior to finalizing thedesign. Several of the initial design parameters were modifiedbased on the above test results and additional observations duringtesting. Changes were made to the operational temperature range(changed to -40 to 85[degrees]C) and probe density (ideally lessthan 2000 kg/m^sup 3^) parameters, among others. It was also seenthat the type of data memory used in the device was important. Thedesign parameter modifications are presented in Table 3. Based onthese modifications, the commercial device was deemed preferableover the conceptual design presented in this paper (Table 4presents a comparison between the two devices). The commercialdevice case, a bored out aluminum rod with a screw cap and o-ringon one end, was simpler, had a smaller part count, was easier toopen and close, and was stronger at the cylinder ends because therewere no joints between parts. Also, in the interest of dataretention, the commercial logger circuitry, with non-volatilememory, is preferred to the DS2422-based circuit, which loses datawhen power to the circuit is lost. The commercial device also hasfour times more memory capacity than the DS2422-based device.
The impact of the smaller case size of the commercial device wouldhave to be considered, however. Probe recovery would likely beunaffected, since the smallest dimension of the commercialtemperature logger is 26 mm, so capture on a 1 in. (25.4 mm) orsmaller screen would still be possible. These probes are also stilllarge enough to be easily recovered visually in the screen overs.However, a decrease in case size may affect the ability of theprobes to disperse randomly in compost, as they may more easily fitinto pore space than the larger cases tested. If they do fit intothe pore spaces they may have a tendency to move downward sincetheir density is higher than that of the compost. As mentionedpreviously, testing should be done to determine the effect of achange in probe size on dispersion during compost agitation events.
Conclusion
A new method for monitoring temperatures in compost from the pointof view of a random compost particle was desired. To this end, aself-contained temperature logging device was designed, built,tested, and compared with a similar commercially available device.A number of experiments were done to determine i) if the cases andcircuitry would withstand severe impacts, ii) how the density ofthe probes would affect their dispersion within a compost pile,iii) if the devices could respond to a temperature change in asufficiently short period of time, and iv) what probe recoverymethod(s) would be most reliable. Initial testing revealed thataluminum cases may be able to withstand the high-impact procedureof windrow turning, provided that the aluminum is of a sufficientthickness. Probes of densities between 1.8 and 4.4 times thedensity of compost appeared to be distributed randomly duringwindrow turning, as desired. Both the commercial and custom devicehad similar (and adequate) response times to step changes intemperature, and could be easily detected visually during recoveryon a screen.
TABLE 3.
Final design specifications for compost temperature probe.
TABLE 4.
Final design specifications - comparison between tested devices. (1= meets constraint; 0 = does not meet constraint; 1/2 or 3/4 =partially meets constraint or could be adjusted to meet constraint)Constraints are weighted on a scale of 1 to 5.
More testing is recommended to confirm the ability of the devicesto mimic the behaviour of compost particles. Further investigationinto the effects of density on probe behaviour should be carriedout, particularly for devices having the same size as thecommercial logger (as previous testing considered larger devices).It is also important to determine how the temperature monitoringdevices will behave when they reach the edges of compost piles. Theeffectiveness of probe recovery via screening should also beconfirmed for probes which have undergone an extended period ofcomposting, as cohesion between compost and the probes may makethem more difficult to see.
Either of the devices, with some improvements, could be used forthe desired application. The main improvement required for bothdevices is more secure battery mounting (to avoid unnecessary powerloss). The commercial device, however, had two main advantages overthe custom device: i) a superior case design (with anodizedaluminum and solid ends, which made it stronger), and ii) theability to retain data during a power failure. It is thereforerecommended that further testing focus on the commercial device.
Acknowledgements
This project has been undertaken with funding assistance from theNatural Sciences and Engineering Research Council of Canada (NSERC)and the Edmonton Waste Management Centre of Excellence. The authorsthank Curt Stout, Mark Adcerman, Glenn Alloway, Greg Brandon,Skyler Hudson, and Janet Jackson from the University of Alberta'sMechanical Engineering Department for their assistance with andsuggestions regarding the design of the temperature probe housing.They also thank Will Bauer for his insights into the electronicsdesign, and Dr. Andy Knight from the University of Alberta for useof his electronics laboratory.
References
Agnew, J. M., and J. J. Leonard. (2003). The physical properties ofcompost. Compost Sci. Util., 11(3), 238-264.
Avnimelech, Y., R. Eilat, Y. Porat, and P. A. Kottas. (2004).Factors affecting the rate of windrow composting in field studies.Compost Sci. Util., 12(2), 114-118.
Bari, Q. H., and A. Koenig. (2001). Effect of air recirculation andreuse on composting of organic solid waste. Resour. Conserv.Recycl., 33(2), 93-111.
Brannen, J. P., D. M. Garst, and S. Langley. (1975). Inactivationof Ascaris lumbricoides eggs by heat, radiation, andthermoradiation. SAND 75-0163, Sandia Labs, Albuquerque, NewMexico.
Catalano, M. (2006). Personal Communication. Geonics LimitedMississauga, Ontario.
CCME. (2005). Guidelines for Compost Quality. PN1340, CanadianCouncil of Ministers of the Environment, Winnipeg, Minnesota.
Deportes, I., J. L. Benoit-Guyod, D. Zmirou, and M. C. Bouvier.(1998). Microbial disinfection capacity of municipal solid waste(MSW) composting. J. Appl. Microbiol., 85(2), 238-246.
Fernandes, L., W. Zhan, N. K. Patni, and P. Y. Jui. (1994).Temperature distribution and variation in passively aerated staticcompost piles. Bioresour. Technol., 48(3), 257-263.
Fichtner, K., S. Gamble, and A. Yee. (2003). Composting biosolidsusing the Gore Cover. System selection, construction, operation.Proceedings of the 2nd Canadian Organic Residuals RecyclingConference, Penticton, British Columbia, 12 p.
Fischer, J. L., T. Beffa, P. F. Lyon, and M. Aragno. (1998).Aspergillus fumigatus in windrow composting: effect of turningfrequency. Waste Manage. Res., 16(4), 320-329.
Haug, R. T. (1993). The Practical Handbook of Compost Engineering,Lewis Publishers, Boca Raton, Florida.
Hay, J. C. (1996). Pathogen destruction and biosolids composting.BioCycle, 37(6), 67-76.
Hess, T. F., I. Grdzelishvili, H. Q. Sheng, and C. J. Hovde.(2004). Heat inactivation of E-coli during manure composting.Compost Sci. Util., 12(4), 314-322.
Joshua, R. S., B. J. Macauley, and H. J. Mitchell. (1998).Characterization of temperature and oxygen profiles in windrowprocessing systems. Compost Sci. Util., 6(4), 15-28.
Kowal, N. E. (1985). Health Effects of Land Application ofMunicipal Sludge. EPA/600/1-85/015, Health Effects ResearchLaboratory, Office of Research and Development, U.S. EnvironmentalProtection Agency, Research Triangle Park, North Carolina.
Lau, A. K., P. H. Liao, and K. V. Lo. (1993). Evaluation of swinewaste composting in vertical reactors. Journal of EnvironmentalScience and Health, Part A: Environmental Science and Engineering,28(4), 761-777.
Little, M. D. (1980). Agents of health significance: parasites. In:Sludge - Health Risks of Land Application. Edit: G. Bitton, B. L.Damron, G. T. Edds, and J. M. Davidson, Ann Arbor SciencePublishers, Inc., Ann Arbor, Michigan.
Madge Tech. (n.d.). Data Logger and Software Operating Manual -Version 2.00.61. Madge Tech, Warner, NH.
Maxim/Dallas. (2003). Application Note 2810 - DS2422 Trim Procedureand Software Correction. Maxim Integrated Products/DallasSemiconductor.
McCartney, D., and G. Eftoda. (2005). Windrow composting ofmunicipal biosolids in a cold climate. J. Environ. Eng. Sci., 4(5),341-352.
Pichtel, J. (2005). Composting MSW. In: Waste Management Practices:Municipal, Hazardous, and Industrial. CRC Press, Boca Raton,Florida.
Rubin, A. R. (1996). Is time-temperature really important? 10thAnnual Residuals and Biosolids Management Conference: 10 Years ofProgress and a Look Toward the Future, Alexandria, Virginia, 17.51-17.57.
Salter, C., and A. Cuyler. (2003). Pathogen reduction in foodresiduals composting. BioCycle, 44(9), 42-51.
Scott, M. (2005). Personal Communication. Edmonton Waste ManagementCentre of Excellence (EWMCE) Edmonton, Alberta.
Stentiford, E. I. (1996). Composting control: principles andpractice. In: The Science of Composting: European CommissionInternational Symposium. Edit: M. de Bertoldi, P. Sequi, B. Lemmes,and T. Papi, Blackie Academic & Professional, London, UnitedKingdom.
Strader, R., and G. S. Bromhal. (1997). Evaluation of temperaturefluctuations in a windrow composting operation. Proceedings of the1997 Air & Waste Management Association's 90th Annual Meeting &Exhibition, Jun 8-13 1997, Toronto, Canada, 10 p.
TI Support. (2005). Personal Communication. Texas InstrumentsSemiconductor Technical Support.
TMECC. (2001). Test Methods for the Examination of Composting andCompost, The US Composting Council Research and EducationFoundation, and the United States Department of Agriculture.Turner, C, A. Williams, R. White, and R. Tillett. (2005). Inferringpathogen inactivation from the surface temperatures of compostheaps. Bioresour. Technol., 96(5), 521-529.
USEPA. (1999). Control of pathogens and vector attraction in sewagesludge. EPA/625/R-92/013 Revised October 1999, United StatesEnvironmental Protection Agency, Office of Research andDevelopment, National Risk Management Laboratory, Center forEnvironmental Research Information, Cincinnati, Ohio.
Van Herk, F. H., T. A. McAllister, C. L. Cockwill, N. Guselle, F.J. Larney, J. J. Miller, and M. E. Olson. (2004). Inactivation ofGiardia cysts and Cryptosporidium oocysts in beef feedlot manure bythermophilic windrow composting. Compost Sci. Util., 12(3), 235-241.
Wichuk, K. M., and D. McCartney. (2007). A review of theeffectiveness of current time-temperature regulations on pathogeninactivation during composting. J. Environ. Eng. Sci., 6(5), 573-586.
Yee, A. (2005). Personal Communication. Edmonton Waste ManagementCentre of Excellence (EWMCE) Edmonton, Alberta.
Yee, A. (2006). Personal Communication. Edmonton Waste ManagementCentre of Excellence (EWMCE) Edmonton, Alberta.
Kristine M. Wichuk and Daryl McCartney
Dept. of Civil and Environmental Engineering, University ofAlberta, Edmonton, Alberta, Canada
Copyright J.G. Press Inc. Spring 2008
(c) 2008 Compost Science & Utilization. Provided by ProQuestInformation and Learning. All rights Reserved.
Source: Compost Science & Utilization
More News in this Category
In order to account for the effects of initial location within thepile on probe movement, the probes were placed in three differentlocations, roughly at the top, middle, and bottom of a windrow atthe co-composter cure site. One third of the probes of each densitybeing tested were placed at each height. All probes were placed inthe same lateral position in order to minimize the search areaafter turning. The heights of the probes in each position weremeasured and recorded. The windrow was then turned once and theprobes recovered. This was accomplished by digging through thecompost by hand, in order not to affect the location of the probesduring their recovery. When a probe case was located, its finalheight was recorded. The second run was carried out approximately 4weeks after the first run, in the same windrow as the first trial.The procedure was similar, with the exception of the number ofcases introduced, the method of securing the end caps to thealuminum tube, and the densities tested. For this trial, 27 casesof each of three densities were introduced into the windrow. Theend caps were secured to the aluminum tubing using electrical tapeonly, as duct tape tended to tear during the first trial. The highdensity was changed to 2000 kg/m^sup 3^, which is slightly higherthan the density of the commercial temperature logger. The lowdensity was kept at 800 kg/m^sup 3^, and thus the midpoint was 1400kg/m^sup 3^. The remainder of the procedure was the same as duringthe first trial, with one third of the probes of each density beingplaced at the top, middle, and bottom of the windrow, the windrowbeing turned, and the location of the probes after turning beingrecorded.
During the second run, the wet bulk density of the compost wastested, in order to have a point of comparison between probe andcompost densities. Compost was collected from four locations (top,bottom, and two middle locations) in the windrow used for testing(after turning), in the same zone as the temperature probe caseswere placed. These four samples were amalgamated in a 5-gallonbucket with a lid and stored for 10 days prior to analysis. Theprocedure used for the bulk density determination was amodification of TMECC Method 03.03-A/03.01-A (steps 10.3 to 10.5only) (TMECC 2001).
Change in height data (i.e. the difference between initial andfinal heights) was analyzed using a two-factor analysis of variance(ANOVA), with density and initial height being the factorsconsidered. The effect of density on probe movement was of majorconcern, but the initial height of the probes also needed to beconsidered because the starting position affected the possibledirections (up or down) and magnitudes of the height change. TheANOVA analysis provided insight into whether probe densitysignificantly affected the change in height of the probes afterturning.
The cross-sectional dimensions of the windrow were measured (i.e.width at the top, w^sub 1^, width at the bottom, w^sub 2^, andheight, h) (see Figure 1). From these dimensions, it was possibleto approximate the distance, d, above ground at which an equalvolume of compost lay above and below. Since probes were initiallyplaced in equal numbers at the top, middle, and bottom of the pile,the corresponding change in height, Deltah, to this volume midpointshould be the difference between this distance, d, and the half thepile height (i.e. the mean initial location). It was postulatedthat random probe movement during windrow turning would beindicated by a mean change in height corresponding to the volumemidpoint, since there would be a 50% chance that material would layabove or below this line. A t-test was used to evaluate this.
FIGURE 1. Approximation of a windrow cross-section; dimension "d"indicates the level in the pile where the volume of material aboveand below the line is approximately equal
Robustness of Circuitry
It was also of prime importance that the temperature loggingcircuit inside the housing remain intact and functioning for theduration of the data-logging mission, regardless of the conditionsencountered. Thus, during tests of density and housing materialstrength in the co-composter cure site windrow, some quickexperiments were also done to examine the robustness of theelectronics to severe impacts (as occur during windrow turning).
During the first run, an operating commercial temperature loggerwas introduced into the middle of the windrow, as was anoperational DS2422-based temperature logging circuit with a singleAA battery. The DS2422 circuit was enclosed in a section ofaluminum tubing in such a way that minimal movement of the circuitboard would occur inside the housing, and the tube was capped witha 50.8 mm diameter, 1.58 mm thick polycarbonate circle on each end.The windrow was turned, the circuits recovered, and the electronicsexamined for damage.
During the second windrow-turning trial, in addition to the twotemperature logging circuits listed above, six cases containing 1/2AA sized lithium batteries soldered onto a circuit board wereplaced into the windrow at the bottom. Three of the cases containedtabbed batteries (TL-5902/T) and the other three containedbatteries with solder pins (TL-5902/P). Again, the circuits wererecovered after windrow turning and examined for damage.
Temperature Response Time
The response time of a temperature sensor is dependent in a numberof factors, including the thermal conductivity of the sensor itselfand of the materials (such as the case walls and air space) betweenthe sensor and the temperature of interest. It was desired that theamount of time required for both the commercial and DS2422- basedloggers to respond to a step change in temperature be determined.An excessively long response time would be undesirable becausetemperature fluctuations of interest may be overlooked; three hoursor less was set as the target.
A two-point temperature calibration (at approximate temperatures of-10[degrees]C and 60[degrees]C) was done for each device, to ensurethat both would provide the same data under equivalent conditions.The two-point calibration was completed as specified in thedocumentation for each device (Madge Tech n.d.; Maxim/Dallas 2003).After calibration, each device was set up to record temperatureonce every 60 s and enclosed in its housing (the DS2422 circuithousing was as described in section 4.4.3, except that the end capswere 4.76 mm thick). The temperature response times of both deviceswere evaluated over temperature changes of approximately +-10[degrees]C, +-15[degrees]C, +-20[degrees]C, +-30[degrees]C, and+- 40[degrees]C. These changes were achieved by moving the probesbetween a room temperature zone and a pre-heated lab oven. When thedevices were moved from one temperature zone to another, they wereallowed to equilibrate for a period of at least 2 hours (based onpreliminary observations), to ensure that sufficient time had beenallowed for the temperature readings to stabilize. The timerequired for a given temperature rise or fall was evaluated whenthe device had attained a temperature within 0.5[degrees]C of itspeak or minimum temperature, respectively.
After all specified positive and negative temperature change cycleshad been completed, temperature data was downloaded from eachdevice and evaluated to determine the response time for each stepchange in temperature.
This test was performed at the University of Alberta.
Probe Recovery
It was desired that as close as possible to 100% of the probesintroduced into a composting system be recovered, so as not to losecollected data. Several probe detection and recovery methods wereconsidered during the design phases. Of those considered, visualrecovery during screening and location by metal detection wereconsidered the most viable options.
The efficiency of probe recovery during screening was tested byintroducing 30 sections (six each of the following colors: gold,blue, yellow, orange, and unpainted) of aluminum tubing into arotary drum screener along with a loader bucket of compost. Theovers of the screen were scanned visually for the aluminum tubesand all those spotted were collected and counted. The differentcolour tubes were evaluated for relative visibility levels. Thistest was repeated twice: once in the finished product of theGore(R) composter and once in compost from the co-composter.
Two trials were also done to determine how easily the temperatureprobes could be located using metal detection. First, the ease oflocating the aluminum sections with a relatively basic, low-costhand-held metal detector (Garret Treasure Ace 100) was tested. Themaximum depth that the aluminum probes could be sensed with themetal detector was determined by burying a piece of aluminum tubingin screened biosolids compost at a known location at a known depth,varying from 1 cm to 20 cm. The metal detector was set on its mostsensitive setting and the compost was scanned. The depth at whichthe piece of aluminum could no longer be sensed was noted. It wasalso of interest to see how easy it would be to locate a stainlesssteel probe (in case the housing material was modified) and to seehow the performance of an industrial detector would compare withthe hobby model used in the first test. Thus, a magnetic "pinfinder" (Schonstedt GA-52Cx Magnetic Locator) was used to locatemagnetic stainless steel rods buried in a co-composter cure pileand also in a yard waste pile. A similar test to that done with theGarrett Treasure Ace was performed.
Effect of Magnetic Field on Data Memory
In some composting operations, magnetic separation is employedeither at the front-end or during screening of the final product.Such operations use strong magnetic fields to separate out andcapture ferrous materials. It is therefore possible that thetemperature loggers may enter into a relatively strong magneticfield during their use. It was of interest to determine whateffect, if any, a strong magnetic field would have on a device, asdata loss or any change in functioning would be undesirable. Theeffect of magnetic fields is not a standard test for mostsemiconductor devices (TI Support 2005), and was therefore unknown.
An experiment was conducted on both the commercial device and theDS2422-based circuit. A temperature-logging mission was started oneach device, and data was collected for one day. This data wasdownloaded and saved without stopping the data-logging missions orerasing data. The devices were then enclosed in their cases and putunder the magnetic separation unit at the EWMCE's co-composter fortwo minutes (which exceeds the amount of time it would normallytake for an object to pass under the magnet). Data was downloadedfrom each device and compared to the data downloaded prior tomagnet exposure. A second temperature-logging mission was startedon both devices and the data from that mission examined to ensurethat device functioning was not affected. Results
Material Strength and Robustness
None of the 16, 14, or 11-gauge aluminum tubes suffered any damage,aside from minor paint chips, when they were dropped from a loaderbucket onto a packed gravel surface. When driven over by afront-end loader the 14- and 16-ga. tubes were crushed, while the11- ga. tube was compressed only slightly (Figure 2). It isimportant to note that this experiment was conducted without therecommended end caps in place, which may have made the casesstronger by providing structural support at the ends.
During the windrow-turning tests, aluminum sections of all gaugeswere dented by the turner's auger blades. Some of the most severedamage can be seen in Figure 3. While all gauges of aluminum wereaffected to some degree, the most acute damage occurred to the16-ga. tubing and the least amount of severe denting occurred to the11- ga. tubes. More and larger dents tended to occur at the ends ofthe tubes rather than in the middle. Again, it should be noted thathaving the recommended end caps in place may have provided someadditional structural support and hence reduced the severity of thedenting at the tube ends.
FIGURE 2. Damage to aluminum tubes driven over by a front-endloader. a) 16-ga. tube. b) 14-ga. tube. c) 11-ga. tube.
FIGURE 3. Examples of damage to aluminum tubes after windrowturning, including denting and deformation. a) 16 ga. tubes. b) 14-ga. tubes. c) 11-ga. tubes.
FIGURE 4. Chemical and microbiological degradation of aluminumhousings recovered after four weeks in a compost cure pile (darkspots indicate areas of damage): a) unpainted, unanodized tube andb) anodized commercial logger case.
Qualitative observations of damage sustained to the aluminum casesdue to chemical and biological factors were also made during thewindrow test. Some of the painted cases darkened in colour afteronly a few days in the cure pile. The gold cases were beginning toshow signs of corrosion or microbial degradation of the paint andaluminum after having been in compost for less than four days. Oneunpainted case was recovered during the second trial after spendingnearly a month in the windrow; this case was severely pitted innumerous locations due to some combination of corrosion andmicrobial degradation. The housing of the commercial temperaturelogger, made from the same aluminum grade, but anodized, was notaffected by chemical or microbiological stresses after the sameamount of time in the cure pile (see Figure 4).
Probe Density
During the first trial, a significant amount of data was losteither because the probe housing was not recovered or because oneor both ends of the case fell off, resulting in a significantdensity change. Out of 26 housings introduced to the windrow, only20 were recovered. Of these, 9 had been broken during turning andsuffered a change in density, leaving only 11 probes intact. Thus,only 11 probes provided useful information for data analysis. Atwo-factor ANOVA was performed on these 11 data points for changein height data. The results of the analysis indicated that, oncethe initial location of the probes was taken into account, therewas not a significant difference in height changes due to differentdensities. However, since there was only a small amount of usefuldata obtained during the first trial, no strong conclusions couldbe drawn regarding the affect of density on probe movement.
A second trial was performed because so many probes were lost ordamaged during the first trial. Three times as many probe housingswere used and the end caps better secured. During the secondexperiment 7 probes were not recovered and two broke. However, thelarge amount of remaining data meant that a more meaningfulanalysis could be completed than was possible after the firsttrial. The dimensions of a cross-section of the windrow weremeasured (as in Figure 1). The pile height, h, was around 1.5 m.The width of the windrow at the top, w^sub 1^, was approximately1.2 m. The width at the bottom of the pile, w^sub 2^, was roughly5.5 m. From these dimensions, it was determined that there wereapproximately equal volumes of material above and below d = 0.53 m,which corresponded to a change in height, Deltah, of -0.22m.Compost bulk density during this trial was 446.9 +- 8.7 kg/m^sup3^.
From the average change in height data for each density, itappeared that there was a slight trend toward more downward probemovement as density increased. Interestingly, a best-fit line(linear regression) through the average change in height datacrossed the volume midpoint (Deltah = -0.22 m) at a density of 452kg/m^sup 3^, which is nearly the same as the compost bulk density.However, when the data was weighted to take into account thedifferent numbers of probes that were lost from each height, thetrend toward downward movement with increasing density was not asobvious (Figure 5) and the volume midpoint was crossed at a densityof 246 kg/m^sup 3^. Additionally, after accounting for the effectsof initial probe location, an ANOVA analysis (95% confidence level)indicated no significant difference in change in height databetween probes of different densities. In other words, the ANOVAresults indicate that varying probe density from 800 kg/m^sup 3^ to2000 kg/ m^sup 3^ appeared not to significantly affect probemovement.
FIGURE 5. Plot of change in height data after windrow turning (raw,averaged, and weighted average) for probes of different densities.
T-tests indicated that for all densities, neither the average northe weighted average change in height significantly differed fromthe compost volume midpoint. This result, along with the ANOVAresult that there was no significant difference in change of heightbetween different densities, seems to indicate that there wasrandom dispersion for probes with densities between 800 and 2000kg/m^sup 3^ during turning of a windrow with a bulk density near450 kg/ m^sup 3^.
Robustness of Circuitry
During the first trial, only the DS2422 device was recovered afterwindrow turning. All circuit components and solder contactsremained intact during windrow turning. However, one of the tabstore off of the battery, so that power to the circuit was lost (seeFigure 6). When the battery was replaced the circuit functionedproperly; however, no data was retained. During the second trial,several 1/2AA-sized batteries of both tabbed and solder pinvarieties were tested in order to see if the battery problemsencountered during initial testing would occur again, and to see ifpins would be stronger than tabs. After turning, the commerciallogger was recovered along with the DS2422 circuit, all three casescontaining tabbed 1/2AA batteries, and one of the three casescontaining batteries with solder pins. All of the solderedbatteries had one or both tabs/pins broken. Again, the onlyapparent problem with the DS2422 circuit was the broken battery,and the device functioned normally when it was replaced.
FIGURE 6. Damage to the DS2422-based temperature logging circuitafter being turned once by a windrow turner. One of the batterytabs was torn off.
The battery in the commercial temperature logging device wassimilar to the other 1/2AA batteries with pins that were tested,except that it was placed in a socket rather than being soldereddirectly into the circuit. Sometime during the three windrowturning activities experienced by the commercial logger (one duringthe initial trial, one during the second trial, and one betweentrials), one of the battery pins came out of its socket and thedevice lost power. Another main component of the circuit, amicrocontroller, also popped out of its socket, and a diode wasbroken in half (see Figure 7). It was hypothesized that the diodebroke because it was hit when the battery came out of its socket.All other components of the commercially available device weresoldered onto the circuit board and survived three passes of thewindrow turner. When all broken or loose parts were replaced, thelog-ger resumed normal operation. Data memory was full, indicatingthat damage to the device likely occurred during the third turningevent.
Figure 7. Damage to the internal circuitry of the commercialtemperature logger after being turned three times with a windrowturner. One of the components came completely out of its socket.
TABLE 2.
Time required for each of the commercial and DS2422-based devicesto respond to different changes in temperature.
Temperature Response Time
The results of the temperature response test are presented in Table2. The commercial and DS2422-based temperature loggers had similarresponse times. The maximum amount of time needed for either deviceto respond to a step change in temperature was 81 minutes, whicheasily met the desired three-hour limit set on response time.Though the data shows an overall trend of longer response times forlarger changes in temperature, the temperature probes are notexpected to experience a step change in temperature much more thanthe largest change tested. Therefore, the temperature responsetimes of both devices were deemed adequate.
Probe Recovery
When various colours of aluminum tubing were screened with bothbiosolids and mixed biosolids/ MSW compost, a person visuallyscanning the screen overs easily recovered 100% of the tubing. Inthe biosolids compost, all colours were equally easy to see.However, in the mixed compost the gold and unpainted cases tendedto blend slightly more than the other colours with non-compostablematerials (such as bits of coloured plastic, glass, and metal) inthe product. Nonetheless, it appeared that screening the probes outof finished compost would be a viable method of recovery. The low-cost metal detector used to evaluate the recovery of aluminum probecases performed quite poorly. A case placed at a known locationcould only be detected up to a depth of 15 cm. Numerous otherobjects were also picked up by the metal detector (even in screenedbiosolids compost). In fact, other objects were picked up with suchfrequency that an operator would be unable to determine whether thedetector was signalling the presence of a probe or of someextraneous object. Reducing the sensitivity of the detector was noteffective in reducing this interference. Additionally, with reducedsensitivity the maximum detection depth decreases.
The magnetic locator was also found to be ineffective for use inprobe recovery. In the co-composter cure pile, a stainless steelbar was easily detected, but other objects in the pile were alsopicked up. As was the case with the metal detector used to locatealuminum tubing, it would be virtually impossible to distinguishbetween a stainless steel probe and other ferrous objects in thecompost. The situation was even worse in the yard waste windrow,with the detector giving strong signals whether or not thestainless steel bars were in the pile.
Effect of Magnetic Field on Data Memory
After exposure to a magnetic field generated by a ferrous magneticseparation unit, temperature data collected by the DS2422 deviceand the commercial device was compared to that collected prior tothe exposure, and was found to be the same. Data collectioncontinued, unaffected, during and after subjection to a magneticfield. The devices also both operated normally when they werestopped and started again, indicating that a strong magnetic fieldhad no effect on the operation of either device.
Discussion and Recommendations
Testing of the designed temperature logging device and a similarcommercially available model revealed that, while both devicesshowed potential as far as performing their desired function,neither device was ideal. Improvements could be made to bothdevices in order to obtain a better fit between the specifiedparameters and the actual performance of the probes.
One of the most important features of the temperature probe was anability for the devices to mimic the behaviour of random particlesof compost during agitation and settling. The average position of aprobe should tend toward the volume midpoint of a compost pile(i.e. there should be a random distribution of probes around thelevel at which there are equal volumes of compost in the top andbottom of the pile). Three probe densities were tested (800,1400,and 2000 kg/m^sup 3^) in compost with a bulk density ofapproximately 450 kg/m^sup 3^. An ANOVA analysis indicated nosignificant differences between the different probe densities, andthere were no statistically significant difference between theaverage change in height data and the volume midpoint for all probedensities. These results imply that random probe distribution wasachieved. However, they were based on a single experiment;additional trials are recommended to confirm the results. It shouldalso be noted that even though random distribution was observedduring this trial, the same results may not be seen with probes ofdifferent sizes or in a pile with fresh feedstock materials. Thisis because during this test the probe size was significantly largerthan the average pore size within the compost heap and thus theprobes were easily held up by the compost matrix. If the probe sizeapproaches the size of a compost pore, it will likely become easierfor dense probes to move downward in the pile. Hence, if the probedimensions are decreased such that they become similar to poresize, this test should be repeated.
Edge effects should also be considered in future trials todetermine how well the probes are able to mimic random probebehaviour. Such effects may include an inability for a windrowturner to pick up a small probe sitting right at the ground (thoughin this case the probe would be able to monitor conditions seen bya compost particle sitting at ground level). Also, since the probesare cylindrical, if they reach the edge of the pile there is a highprobability that they will roll off the pile, whereas a compostparticle may not. This is an undesirable situation, as valuabledata could be lost. A possible way to prevent rolling would be toattach a counterweight or hook to the device.
It was also important to be able to recover all (or nearly all) ofthe probes introduced into the compost. Screening appeared to be afeasible method of probe recovery. However, it is recommended thatadditional screening tests be performed after probes have undergonea complete compost cycle, as it is possible that the compression ofmaterials within a pile could cause compost to stick to the casesand thus make them difficult to see. The results reported in thispaper refer to a screening test done immediately after adding casesto relatively dry compost. Two attempts at probe recovery by metaldetection failed, partly because detection depth was small andpartly because there was interference from extraneous objects inthe compost pile. It is possible that a more expensive detector mayperform better; some higher end detectors have visual displayswhich allow specific objects to be identified by signal/wave shape.These detectors are also able to penetrate deeper, though maximumdepths of only a few feet are reported, which may still be tooshallow (Catalano 2006). At this point, screening seems to be themost viable probe recovery method. Since neither device appeared tobe affected at all by exposure to a strong magnetic field, it isnot of concern if screening operations employ magnets to removeferrous materials from the compost.
The temperature response times of both the commercial and DS2422-based temperature loggers were satisfactory; the response timeswere short enough that it is unlikely any significant variation intemperature (with time) would be overlooked. However, if majorchanges are made to the probe design, these tests should berepeated to confirm that the results are still adequate. Changes tothe amount of air space inside the probe or changes to thematerials used could potentially affect the amount of time requiredfor the device to respond to a temperature change.
As far as the probe housing is concerned, it was seen that 16-gauge aluminum was not sturdy enough to survive relativelyunscathed during activities such as windrow turning. It isrecommended that a thicker gauge, such as 14 or 11, be used in thefinal design. The 11- gauge tubing did suffer some denting duringturning, but the severity of the dents was small relative to thoseon the other two gauges. However, since 11-gauge adds extra mass tothe probes 14- gauge might be a better choice since it survivedreasonably well during high impact situations and would contributeless to increasing probe density. It was seen that dent frequencyand severity was greater at the ends of all gauges of aluminumtubes than in the middle. Severe denting at the tube ends is causefor concern, given the proposed design of the case (an aluminumtube with polypropylene end caps fitting inside the tube); it ispossible that even minor dents at the tube ends tubes may damagethe seal between the end caps and the aluminum tube. This couldlead to exposure of the electronic circuitry to moisture, compost,and/or contaminants. It is particularly important to preventmoisture from coming into contact with the lithium battery used, asthis may present an explosion hazard. It should be noted, however,that had the suggested end caps been in place during testing, theremay have been additional structural support and dent severity mayhave been reduced.
A brightly coloured case (i.e. blue, yellow, or orange) isrecommended to increase visibility. Aluminum can be anodizeddifferent colours; this is advised, as anodizing seemed to beeffective in preventing corrosion and would limit paint chippingand discoloration.
The most serious problem encountered during windrow turningexperiments was that electronic components and batteries placed insockets popped out of the sockets, and batteries soldered to thecircuit board suffered broken pins or tabs. The ability of thedevice to function as desired was affected when power was lost and/or circuit components failed. The use of solder-mount chips ratherthan sockets would solve half the problem. However, anothersolution is needed for the problem of battery breakage.
It is hypothesized that battery breakage occurred due toconservation of momentum; when the probe case was accelerated ordecelerated, the battery on the inside of the case did notexperience the same change in momentum. In other words, when thecase, and hence the circuit board, was stopped, the battery wasable to keep moving and the solder tabs broke. This was a problemfor the battery but not for other soldered parts because thebattery was relatively large and heavy (momentum is a function ofmass) and because the battery was attached to the circuit board inonly two places (at either end), while other parts were attached atseveral points along two edges. It is important to find a solutionto the problem of battery breakage, as this causes power loss andhence data loss. The solution is probably as simple as finding abetter way to secure the battery to the circuit board. One optionwould be to create a mould that would fit snugly around the circuitboard and battery, preventing movement of the internal parts.Another option, suggested by the manufacturer of the commercialdata logger, would be to wrap tape around the battery to keep it inplace; electrical tape may be suitable for this purpose, as it wasstrong enough to hold end caps in place during windrow turning. Thetape option may be preferable because the temperature response timemay be increased by the mould material. Tape would also add lessweight to the probe. Testing is recommended prior to finalizing thedesign. Several of the initial design parameters were modifiedbased on the above test results and additional observations duringtesting. Changes were made to the operational temperature range(changed to -40 to 85[degrees]C) and probe density (ideally lessthan 2000 kg/m^sup 3^) parameters, among others. It was also seenthat the type of data memory used in the device was important. Thedesign parameter modifications are presented in Table 3. Based onthese modifications, the commercial device was deemed preferableover the conceptual design presented in this paper (Table 4presents a comparison between the two devices). The commercialdevice case, a bored out aluminum rod with a screw cap and o-ringon one end, was simpler, had a smaller part count, was easier toopen and close, and was stronger at the cylinder ends because therewere no joints between parts. Also, in the interest of dataretention, the commercial logger circuitry, with non-volatilememory, is preferred to the DS2422-based circuit, which loses datawhen power to the circuit is lost. The commercial device also hasfour times more memory capacity than the DS2422-based device.
The impact of the smaller case size of the commercial device wouldhave to be considered, however. Probe recovery would likely beunaffected, since the smallest dimension of the commercialtemperature logger is 26 mm, so capture on a 1 in. (25.4 mm) orsmaller screen would still be possible. These probes are also stilllarge enough to be easily recovered visually in the screen overs.However, a decrease in case size may affect the ability of theprobes to disperse randomly in compost, as they may more easily fitinto pore space than the larger cases tested. If they do fit intothe pore spaces they may have a tendency to move downward sincetheir density is higher than that of the compost. As mentionedpreviously, testing should be done to determine the effect of achange in probe size on dispersion during compost agitation events.
Conclusion
A new method for monitoring temperatures in compost from the pointof view of a random compost particle was desired. To this end, aself-contained temperature logging device was designed, built,tested, and compared with a similar commercially available device.A number of experiments were done to determine i) if the cases andcircuitry would withstand severe impacts, ii) how the density ofthe probes would affect their dispersion within a compost pile,iii) if the devices could respond to a temperature change in asufficiently short period of time, and iv) what probe recoverymethod(s) would be most reliable. Initial testing revealed thataluminum cases may be able to withstand the high-impact procedureof windrow turning, provided that the aluminum is of a sufficientthickness. Probes of densities between 1.8 and 4.4 times thedensity of compost appeared to be distributed randomly duringwindrow turning, as desired. Both the commercial and custom devicehad similar (and adequate) response times to step changes intemperature, and could be easily detected visually during recoveryon a screen.
TABLE 3.
Final design specifications for compost temperature probe.
TABLE 4.
Final design specifications - comparison between tested devices. (1= meets constraint; 0 = does not meet constraint; 1/2 or 3/4 =partially meets constraint or could be adjusted to meet constraint)Constraints are weighted on a scale of 1 to 5.
More testing is recommended to confirm the ability of the devicesto mimic the behaviour of compost particles. Further investigationinto the effects of density on probe behaviour should be carriedout, particularly for devices having the same size as thecommercial logger (as previous testing considered larger devices).It is also important to determine how the temperature monitoringdevices will behave when they reach the edges of compost piles. Theeffectiveness of probe recovery via screening should also beconfirmed for probes which have undergone an extended period ofcomposting, as cohesion between compost and the probes may makethem more difficult to see.
Either of the devices, with some improvements, could be used forthe desired application. The main improvement required for bothdevices is more secure battery mounting (to avoid unnecessary powerloss). The commercial device, however, had two main advantages overthe custom device: i) a superior case design (with anodizedaluminum and solid ends, which made it stronger), and ii) theability to retain data during a power failure. It is thereforerecommended that further testing focus on the commercial device.
Acknowledgements
This project has been undertaken with funding assistance from theNatural Sciences and Engineering Research Council of Canada (NSERC)and the Edmonton Waste Management Centre of Excellence. The authorsthank Curt Stout, Mark Adcerman, Glenn Alloway, Greg Brandon,Skyler Hudson, and Janet Jackson from the University of Alberta'sMechanical Engineering Department for their assistance with andsuggestions regarding the design of the temperature probe housing.They also thank Will Bauer for his insights into the electronicsdesign, and Dr. Andy Knight from the University of Alberta for useof his electronics laboratory.
References
Agnew, J. M., and J. J. Leonard. (2003). The physical properties ofcompost. Compost Sci. Util., 11(3), 238-264.
Avnimelech, Y., R. Eilat, Y. Porat, and P. A. Kottas. (2004).Factors affecting the rate of windrow composting in field studies.Compost Sci. Util., 12(2), 114-118.
Bari, Q. H., and A. Koenig. (2001). Effect of air recirculation andreuse on composting of organic solid waste. Resour. Conserv.Recycl., 33(2), 93-111.
Brannen, J. P., D. M. Garst, and S. Langley. (1975). Inactivationof Ascaris lumbricoides eggs by heat, radiation, andthermoradiation. SAND 75-0163, Sandia Labs, Albuquerque, NewMexico.
Catalano, M. (2006). Personal Communication. Geonics LimitedMississauga, Ontario.
CCME. (2005). Guidelines for Compost Quality. PN1340, CanadianCouncil of Ministers of the Environment, Winnipeg, Minnesota.
Deportes, I., J. L. Benoit-Guyod, D. Zmirou, and M. C. Bouvier.(1998). Microbial disinfection capacity of municipal solid waste(MSW) composting. J. Appl. Microbiol., 85(2), 238-246.
Fernandes, L., W. Zhan, N. K. Patni, and P. Y. Jui. (1994).Temperature distribution and variation in passively aerated staticcompost piles. Bioresour. Technol., 48(3), 257-263.
Fichtner, K., S. Gamble, and A. Yee. (2003). Composting biosolidsusing the Gore Cover. System selection, construction, operation.Proceedings of the 2nd Canadian Organic Residuals RecyclingConference, Penticton, British Columbia, 12 p.
Fischer, J. L., T. Beffa, P. F. Lyon, and M. Aragno. (1998).Aspergillus fumigatus in windrow composting: effect of turningfrequency. Waste Manage. Res., 16(4), 320-329.
Haug, R. T. (1993). The Practical Handbook of Compost Engineering,Lewis Publishers, Boca Raton, Florida.
Hay, J. C. (1996). Pathogen destruction and biosolids composting.BioCycle, 37(6), 67-76.
Hess, T. F., I. Grdzelishvili, H. Q. Sheng, and C. J. Hovde.(2004). Heat inactivation of E-coli during manure composting.Compost Sci. Util., 12(4), 314-322.
Joshua, R. S., B. J. Macauley, and H. J. Mitchell. (1998).Characterization of temperature and oxygen profiles in windrowprocessing systems. Compost Sci. Util., 6(4), 15-28.
Kowal, N. E. (1985). Health Effects of Land Application ofMunicipal Sludge. EPA/600/1-85/015, Health Effects ResearchLaboratory, Office of Research and Development, U.S. EnvironmentalProtection Agency, Research Triangle Park, North Carolina.
Lau, A. K., P. H. Liao, and K. V. Lo. (1993). Evaluation of swinewaste composting in vertical reactors. Journal of EnvironmentalScience and Health, Part A: Environmental Science and Engineering,28(4), 761-777.
Little, M. D. (1980). Agents of health significance: parasites. In:Sludge - Health Risks of Land Application. Edit: G. Bitton, B. L.Damron, G. T. Edds, and J. M. Davidson, Ann Arbor SciencePublishers, Inc., Ann Arbor, Michigan.
Madge Tech. (n.d.). Data Logger and Software Operating Manual -Version 2.00.61. Madge Tech, Warner, NH.
Maxim/Dallas. (2003). Application Note 2810 - DS2422 Trim Procedureand Software Correction. Maxim Integrated Products/DallasSemiconductor.
McCartney, D., and G. Eftoda. (2005). Windrow composting ofmunicipal biosolids in a cold climate. J. Environ. Eng. Sci., 4(5),341-352.
Pichtel, J. (2005). Composting MSW. In: Waste Management Practices:Municipal, Hazardous, and Industrial. CRC Press, Boca Raton,Florida.
Rubin, A. R. (1996). Is time-temperature really important? 10thAnnual Residuals and Biosolids Management Conference: 10 Years ofProgress and a Look Toward the Future, Alexandria, Virginia, 17.51-17.57.
Salter, C., and A. Cuyler. (2003). Pathogen reduction in foodresiduals composting. BioCycle, 44(9), 42-51.
Scott, M. (2005). Personal Communication. Edmonton Waste ManagementCentre of Excellence (EWMCE) Edmonton, Alberta.
Stentiford, E. I. (1996). Composting control: principles andpractice. In: The Science of Composting: European CommissionInternational Symposium. Edit: M. de Bertoldi, P. Sequi, B. Lemmes,and T. Papi, Blackie Academic & Professional, London, UnitedKingdom.
Strader, R., and G. S. Bromhal. (1997). Evaluation of temperaturefluctuations in a windrow composting operation. Proceedings of the1997 Air & Waste Management Association's 90th Annual Meeting &Exhibition, Jun 8-13 1997, Toronto, Canada, 10 p.
TI Support. (2005). Personal Communication. Texas InstrumentsSemiconductor Technical Support.
TMECC. (2001). Test Methods for the Examination of Composting andCompost, The US Composting Council Research and EducationFoundation, and the United States Department of Agriculture.Turner, C, A. Williams, R. White, and R. Tillett. (2005). Inferringpathogen inactivation from the surface temperatures of compostheaps. Bioresour. Technol., 96(5), 521-529.
USEPA. (1999). Control of pathogens and vector attraction in sewagesludge. EPA/625/R-92/013 Revised October 1999, United StatesEnvironmental Protection Agency, Office of Research andDevelopment, National Risk Management Laboratory, Center forEnvironmental Research Information, Cincinnati, Ohio.
Van Herk, F. H., T. A. McAllister, C. L. Cockwill, N. Guselle, F.J. Larney, J. J. Miller, and M. E. Olson. (2004). Inactivation ofGiardia cysts and Cryptosporidium oocysts in beef feedlot manure bythermophilic windrow composting. Compost Sci. Util., 12(3), 235-241.
Wichuk, K. M., and D. McCartney. (2007). A review of theeffectiveness of current time-temperature regulations on pathogeninactivation during composting. J. Environ. Eng. Sci., 6(5), 573-586.
Yee, A. (2005). Personal Communication. Edmonton Waste ManagementCentre of Excellence (EWMCE) Edmonton, Alberta.
Yee, A. (2006). Personal Communication. Edmonton Waste ManagementCentre of Excellence (EWMCE) Edmonton, Alberta.
Kristine M. Wichuk and Daryl McCartney
Dept. of Civil and Environmental Engineering, University ofAlberta, Edmonton, Alberta, Canada
Copyright J.G. Press Inc. Spring 2008
(c) 2008 Compost Science & Utilization. Provided by ProQuestInformation and Learning. All rights Reserved.
Source: Compost Science & Utilization
More News in this Category
Related News »
In Focus »
Chemical Restricted
Engaging in concept of environmental protection for the Green Olympics, the chemical industry ..
- U.S. team to provide all Olympic ..
- Investors eye coal-to-oil conversion ..
- Chemical education in need of reform
B2B Keywords:
International market Chinese Importer Wholesale trade Wholesale products World trade Wholesale distributors International trade Foreign trade Wholesale distributor Importers Import export business Sell online Help u sell Global trade How to market a product Online supplier Wholesale product
International market Chinese Importer Wholesale trade Wholesale products World trade Wholesale distributors International trade Foreign trade Wholesale distributor Importers Import export business Sell online Help u sell Global trade How to market a product Online supplier Wholesale product
Feature Products»
Hot Products: A | B | C | D | E | F | G | H | I | J | K | L | M | N | O | P | Q | R | S | T | U | V | W | X | Y | Z | 0-9
-
-
About Us -
Links -
Privacy Policy -
Terms of Use -
Site Map -
Contact Us -
Call Center: 0086-95105561 © 2007 Tootoo.com -
