- Email: [email protected]
Self-heating of dried wastewater sludge
Waste Management 33 (2013) 129–137
Contents lists available at SciVerse ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Self-heating of dried wastewater sludge M. Zerlottin a, D. Refosco a, M. Della Zassa b, A. Biasin b, P. Canu b,⇑ a b
Acque del Chiampo, SpA, Via Ferraretta, 20, 36071 Arzignano (VI), Italy Dept. of Industrial Engineering, University of Padua, Via Gradenigo, 5/a, 35131 Padova, Italy
a r t i c l e
i n f o
Article history: Received 27 February 2012 Accepted 22 August 2012 Available online 6 October 2012 Keywords: Self-heating Self-combustion Thermal runaway Wetting heat Tannery sludge
a b s t r a c t We experimentally studied the occurrence of spontaneous self-heating of sludge after drying, to understand its nature, course and remediation. The sludge originates from primary and biological treatment of both municipal and industrial wastewater, the latter largely dominant (approx. 90% total organic carbon, mainly from local tanneries). Dried sludge is collected into big–bags (approx. 1.5 m3) and landfilled in a dedicated site. After several years of regular operation of the landfill, without any management or environmental issue, indications of local warming emerged, together with smoke and smelling emissions, and local subsidence. During a two year monitoring activity, temperatures locally as high as 80 °C have been detected, 6–10 m deep. Experiments were carried out on large quantities of dried sludge (1 t), monitoring the temperature of the samples over long periods of time (months), aiming to reproduce the spontaneous self-heating, under different conditions, to spot enhancing and damping factors. Results demonstrate that air is a key factor to trigger and modulate the self-heating. Water, in addition to air, supports and emphasizes the heating. Unusual drying operation was found to affect dramatically the self-heating activity, up to spontaneous combustion, while ordinary drying conditions yield a sludge with a moderate self-heating inclination. Temperature values as well as heating time scales suggest that the exothermic process nature is mainly chemical and physical, while microbiological activity might be a co-factor. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Wastewater treatment plants yield a large quantity of sludge in addition to purified water. Disposing of sludge may be a significant issue. Dehydrating and drying processes allow to largely reduce the sludge mass, before landfilling. Questions arise about the chemical and biological stability of dried sludge. In the plant of our interest, quite large indeed, dried sludge is collected into PP, with waterproof PE internal layer, big–bags with a volume of approx. 1.5 m3. Bags are subsequently landfilled in a dedicated site. After several years of regular operation of the landfill, without any management or environmental difficulty, local temperature unusually high have been measured. At the same time, landfill produced smoke and malicious odor. Locally, quite significant subsidence was measured, up to 2 m. After a systematic monitoring campaign, lasting more than two years, temperatures locally as high as 80 °C have been detected, in specific zones, typically 6–10 m deep. Local surface subsidence up to 2 m was measured, over a total depth of approx. 12–14 m. While surprised by the significance of the process going on in the landfill, we started a systematic investigation, beginning with a literature survey on comparable scenarios. The self-heating on ⇑ Corresponding author. Tel./fax: +39 0498275463. E-mail address: [email protected] (P. Canu). 0956-053X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.wasman.2012.08.014
a large scale of different materials (carbonaceous material, MSW, municipal sewage sludge) has been reported (Riley et al., 1987; Moqbel et al., 2010; Escudey et al., 2008). Riley et al. (1987) focused on coal in barges, clearly admitting that exhotermic oxidation reactions take place, stressing the role of air. Moqbel et al. (2010) revert to laboratory to identify the auto-ignition temperature of synthetic MWS; they already observed the key role of oxygen to support chemical oxidation reactions. Escudey et al. (2008) carried out field tests on sewage sludge piles extending several meters, for 20 weeks, confirming that temperature could rise up to 90 °C, never causing self combustion, though reporting that it was observed in landfills in their region. In the case of our dry sludge, the first hypotheses concerns the occurrence of some extraordinary biological activity. Literature reports contrasting opinions about that. Li et al. (2008) explicitly support the quantitative generation of heat by biological processes. Others (Poffet et al., 2008; Gholamifard and Eymard, 2009), claim that bacterial activity requires a longer induction time (days or weeks) and sufficient moisture degree. Finally, Yasuhara et al. (2010) like Fu et al. (2005) definitely exclude any connection between self-heating and the bacterial metabolism. The residual moisture in our case is always less than 15% (typically 10%), to prevent the onset of biological processes. However, loss of containment by a big–bag in the landfill, causes sludge to rehydrate, possibly triggering some biological activity. This activity, while
130
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
initially aerobic, can evolve in anaerobic where oxygen lacks, with emission of biogas. Because of the measured temperatures, up to 80 °C, we hardly believe that biological processes are the principal ones taking place. The aerobic processes are known to operate up to a maximum temperature of 70–75 °C, while the anaerobic processes are not active above 60–65 °C. Furthermore, lot of water to rehydrate the sludge is required to provide a suitable environment for biological activity. The landfill is progressively filled by segregating limited deposition areas with LDPE sheets, to repair from rain, preventing quantitative rehydration to take place. Self-heating is likely to occur because of hydration reactions involving salts and oxides in the inorganic fraction, as discussed by Fu et al. (2005) and Walker, 1967. They measured instantaneous heating of RDF and other solids after the addition of water because of what they called ‘‘wetting heat’’. The sudden heating is not compatible with any sort of growing bacterial activity. They proved that the temperature rose from ambient only after the addition of water. However, the hydration heat, if any exists in our case, can be easily overwhelmed by sensible and latent evaporation heat in case a large amount of water is used, as we experienced in remediation attempts successfully applied to the landfill. We can exclude accidental or intentional ignition (fires, cigarettes, etc.) for two reasons. The dried sludge is not easily ignited and any ignition should develop mostly locally, while heating was observed in different zones within the landfill, with the highest temperatures in the deep of the waste. The most likely cause was suggested by Poffet et al. (2008) who conjectured the presence of pyrophoric substances, able to chemically react with oxygen and water very exothermically. Authors (Li et al., 2008) observed self-heating, sometimes leading to self-combustion, of dried sludge from municipal wastewater treatments. They measured temperatures in excess of 600 °C. We were extremely surprised both by the resemblance of the case described in Poffet et al. (2008) with our experience, notwithstanding the supposed difference in the sludge origin, and the extent of the consequences. Interestingly, they remarked a significant presence of sulfur and iron, which is similar to our sludge, though the origin differs. Moreover, awareness of the possibility of fires development in dried sludge bunkers remains apparently underestimated in the management of solid wastes. It was certainly the case at the time of our landfill design, which occurred much earlier that 2008, when (Poffet et al., 2008) was published, but still a large uncertainty remains about the actual chemistry and likelihood of fire occurrence, in practice. Poffet et al. (2008) carried out further small scale investigations in small isolated vessels (Dewar type), as already suggested by Shea and Hsu (1972) reporting that an oxygen flow causes a quicker and more intense heating of the sludge than an inert flux. They also observed that water supports the heating, already suggesting that exothermic reactions could occur, like hydration of metals transition salts. It is known indeed (Zhao, 2006) that the ferrous sulfide, FeS, largely present in the sludge, can react with oxygen (air) at ambient, temperature, with a significant heat of reaction (up to 2500 kJ/mol, according to literature). As expected, the smaller the granules size, the faster the exothermic reactions develop, because of the higher interphase (gas–solids) surface. Also the environmental temperature affects the maximum temperature that solids can achieve, apparently because of a competition between heating by reactions and cooling by heat transfer to the ambient (with lower ambient temperature, the thermal dispersion increases). Being able to reproduce the self-heating of the sludge under controlled conditions is a prerequisite to understand the chemical and physical mechanism, eventually leading to control the heat release and to prevent run-aways. Here we report about large, big– bag scale investigations highlighting the basic features and magnitude of the self-heating phenomena.
What was studied here could be considered one of different phenomena, including the so called smoldering, not well understood yet, though amenable to dramatic consequences. Definitely, they deserve a scientific investigation. Here we report the beginning of our study, i.e. observations on the large scale.
2. Materials and methods 2.1. Sludge Experiments focused on dried sludge, resulting from a single wastewater treatment plant, mainly received from the local industrial tannery district. Sludge is collected both from the primary settling tank (2/3 of total solids produced in the plant are from raw wastewater) and from floatation (due to bacterial activity). The plant of our interest is quite large (1.5 millions equiv. inhabitants) resulting in a total annual mass flow of sludge suspension, at approx. 95% water, in excess of 420,000 tons/y. It is evident the need of reducing the volume. A first dewatering process raises the solids concentration of sludge to 25–32%, then drying brings it to an average of 89% in ordinary conditions. Still, the average production rate of dried sludge approaches 72 tons/d. Details of dewatering and drying are worth discussing, because they may play a role in determining the final activity of the resulting solids. Dewatering is mechanically achieved by filterpresses, where water-laden sludge is squeezed in several tissue-confined chambers. Before filtering, ferrous chloride and organic flocculant are added to the suspension; the first one binds the sulfide ions as insoluble ferrous sulfide, to limit the emission of H2S in working areas and helps with the latter the dewatering capacity of sludge, improving the flocs aggregation thus facilitating the release of water. Dewatered sludge is conveyed to two storage silos (100 m3 each) through screw conveyors. Subsequent drying is achieved thermally, by different methods. Two equal lines, L1 and L2, each one with a 4 t/h evaporation capacity, operate with direct solids heating by convection, with hot gas produced by a natural gas burner. A low concentration of oxygen (7–12%) is maintained in the drier to reduce the dust explosion hazard. Two additional equal lines, L3 and L4, each one with a 2.6 t/h evaporation capacity, were recently added. These are based on thin film evaporation down hot surfaces, externally heated by heat transfer oil. During construction and start-up of L3 and L4 lines, a support line, L5, with a 2 t/h evaporation capacity, was in operation, and produced some of the tested sludge. This combined both direct and indirect drying methods at the same time, both realized as described above. Eventually, sludge in the form of powder with granules of approx. 2 mm mean size, with 85–92% dry solid content and 0.7 kg/ L apparent density, is stored in big–bags subsequently moved to a dedicated landfill. The average composition is shown in Table 1 for the most significant elements. Note the low concentration of heavy metals except for chromium, largely used in the tannery activities, reaching up to 4%wt of solid substance. Iron (up to 2 wt.%) is due to additions at different treatment stages: as FeCl3 after biological treatment, to remove phosphorous from wastewater and improve clarification, furthermore it is added as FeCl2 before dewatering, as discussed above. Another important figure is the total sulfur content, up to 2 wt.%; compared with the small concentration of sulfate and sulfide, it suggests the presence of elemental sulfur. Sludge has a peculiar odor, changing from dewatered to dried; its further change upon self-heating indicates the occurrence of chemical transformations, to a different extent. Due to its composition and origin, mainly industrial (tannery) wastewaters, sludge must be disposed according to specific
M. Zerlottin et al. / Waste Management 33 (2013) 129–137 Table 1 Average composition of dried sludge, based on solid substance (SS). Analyses
Units
pH Dry solids content (@105 °C) Volatile solids content (@600 °C)
– % %
Total oily substances Total nitrogen (TKN) Total chromium Total sulfur Iron Total hydrocarbon Phosphorus Aluminum Sulfate Silicon Zinc Sulfide Manganese Copper Nickel Arsenic Lead Cadmium
g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg
Quantity 8.5 88.5 74.9
(SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS)
132.82 57.23 35.64 19.89 16.48 13.61 9.71 3.01 2.49 2.03 1.39 0.29 0.19 0.05 0.03 <0.023 0.01 <0.001
hazardous waste management regulations. At the present, a landfill totally dedicated to this waste is used, while technical studies to further reduce its volume, exploiting its energetic value (12– 16 MJ/kg) are under way. Over the decades, the company realized and managed nine landfills of its own property, some still receiving sludge while others reached their maximum capacity. At this point, it is covered with different layers: gravel, clay, HDPE sheets, soil and topsoil to prevent infiltration in the waste mass and kept under control for 30 years or more. At the present, dried sludge is disposed of in a single, dedicated landfill (D9), with a capacity of 366,000 m3 over a depth of 12 m. D9 has a waterproof clay shell, covered by a HDPE geomembrane. A network of perforated pipes collects the rain water to drainage wells. Big–bags are orderly placed in the landfill, as shown in Fig. 1; islands of approx. 600 bags are further segregated by means of LDPE sheets, to protect from UV and other atmospheric agents. Big–bags are made
131
of a PP tissue, with treatments against UV; consequently, they are gas permeable. They are then provided with an additional gas-tight PE internal layer (liner). 2.2. Self-heating experiments Experiments to reproduce and characterize the self-heating process were carried out at the big–bag scale, thus using large quantities of dried sludge (1 t). Most tests are carried out on a single big–bag of sludge, isolated inside an industrial enclosure where environmental conditions are more stable. A few test involved more than one bag, closely packed, to investigate cooperative effects. Measurements involve monitoring the temperature over long periods of time (up to several months). Either a single temperature, approximately in the center of the sludge mass, or several points are monitored, Fig. 2. Occasionally, semiquantitative gas emission measurements and IR images were collected. The sludge is poured into the big–bags after drying. The sludge in each monitored bag was separately analyzed, seeking correlations between different reactivity and composition. We investigated several factors: the drying technology, effects of gas permeation through the bag, aeration before filling, water addition, heat dissipation, and environmental influences. Overall, we monitored the temperature in 75 single big–bags, for a total (cumulated) of 90 months of observation, resulting in very extensive experimental campaign during a period of more that 16 months. In the following we illustrate the most significant results. 3. Results and discussion 3.1. Role of air The first issue that was investigated is the role of the inner PE liner gas-tight. Intuition suggested that it could be a severe limitation for water penetration in case excessively hot solids have to be cooled. The same intuition suggests that the PE inner bag could actually support the heating process, limiting the heat dissipation.
Fig. 1. Overview of the landfill dedicated to dried sludge (D9).
132
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
0.4m 0.75m 1.15m 1.25m
0.25m 0.5m 1m
Fig. 2. Schematic of a single big–bag with thermocouples positions.
Fig. 3. Increase of temperature in two big–bags, with and without internal liner.
Two bags, with and without inner liner, were filled with the same dry sludge, placed in a closed space and the central temperature was monitored. Results are shown in Fig. 3. Up to the 7th day, bags where kept closed, as usual in the landfill. The temperature at the end of filling was approx. 50 °C for both samples; in both cases we observed a sudden initial rise of temperature of approx. 10 °C. Surprisingly, the behavior afterwards shows that the inner PE layer kept the temperature rise fairly limited, below approx. 63 °C, while the temperature of the sludge contained in the knitted PP bag rose significantly in a few days, approaching 90 °C. Maximum temperatures are also achieved after quite different time delays, 3–4 days without internal layer, and approx. 1 day with the gas-tight liner. After peaking, temperature gradually decreases, again at a different rate. The internal liner causes a slower cooling. Complete cooling down to ambient temperature took almost 3 weeks for the single big–bag with the simple PP envelope.
The different behavior suggests that contact with air (filling the bag by pouring, having a gas permeable bag) could result in enhancement of the reactivity. To further confirm this tentative conclusion, the bag with PE internal liner was opened after clear, irreversible cooling took place, i.e. at the 7th day. The bag was opened at the top and several small holes drilled on the bag sides to allow permeation of air as much as possible, approaching the big–bag without liner. As clear form Fig. 3, reactivity soon increased enough to temporally overcome the heat dissipation, causing the declining temperature to rise again, up to a secondary maximum. Following the intuition provided by the above experiments, we investigated the effect of a prolonged contact with air before filling the bags. We carried out three tests in parallel, on the same sludge. In one case the standard, 15 min. filling time was applied. The other test achieve a longer filling time, by
133
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
Test a 15min
Test b 35min
Test c 60min
90
Temperature°C
80
70
60
50
40
30
20
0
1
2
3
4
5
6
7
8
Time (d) Fig. 4. Effect of pre-aeration before packaging the sludge into big–bags without inner liner.
100
Temperature°C
90
80
70
60
50
40
30 0
1
2
3
4
5
Time (d) Direct
Mix
Indirect
Fig. 5. Comparison between different shelf-heating in big–bags without inner liner of sludge dried with different technologies.
introducing several pauses, given that the discharge flow rate cannot be modified. Results are shown in Fig. 4. The shape of the profiles are almost identical, but the extent of pre-aeration affects the initial temperature. Moreover, we observed the double, diverging role of air causing both heat dissipation and self-heating enhancement. Results of Fig. 4 suggest that a more thorough exposure to air during filling better dissipate the residual heat content in the sludge, but it is not sufficient to modify the subsequent self heating behavior, which evolve over a much longer time scale and is determined by the capacity to aerate the sludge during that interval (i.e. having a gas permeable or tight containment, as shown in Fig. 3). Thermal behavior of sludge after these additional tests is clearly characterized by its contact with air.
With the availability of different drying technologies, such as direct (L1, L2), indirect (L3, L4) and mixed (L5) heat exchange, expected to expose the sludge to different temperature histories, we planned a comparative test, in parallel on the same dehydrated sludge. Results are shown in Fig. 5. Self-heating of sludge, when in the final bag, is not significantly depending on the drying technology, being direct or indirect heat transfer. A different behavior is observed with sludge from L5, that uses a mixture of both direct and indirect heating (curve labeled Mix in the figure); a lowering of the initial temperature is observed. Because of its structure, L5 discharges the dried sludge through a screw conveyor, water cooled, with external air draft; that allows the sludge to cool and maybe partially react before pouring it into the bag. Eventually, we concluded that the drying method does not affect
134
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
moisture conc. 13,3%
moisture conc. 29,5%
95 85
Temperature (°C)
75 65 55 45 35 25 15 0
5
10
15
20
25
30
35
40
45
Time (d) Fig. 6. Effect of residual moisture in the dried sludge.
95
85
Temperature (°C)
75
addition of water
65
55
45
35
25
15
Time (d) Fig. 7. Reactivation of the self-heating by subsequent additions of water.
significantly conditions.
the
self-heating
behavior,
in
ordinary
drying
3.2. Role of water Furthermore, based on intuition, we expected the self-heating to be stronger with drier solids or, seen on the other side, a larger amount of moisture to mitigate the reactivity. We check this hypothesis comparing the behavior of the same dehydrated sludge that was dried to a different extent, in the same unit (L5). Both were put in gas permeable (no internal PE liner) big–bags. Fig. 6 shows the results, clearly contradicting the presumption. While residual moisture dramatically affects the temperature increase
and intensity of self-heating, the direction was the opposite than expected. If the sludge with a residual moisture slightly above the ordinary (i.e. 13.3% instead of the average of 11%) compares well with previous results, the sludge with a remarkably larger residual moisture (29.5%) shows a much stronger self-heating effect, both in terms of maximum temperature achieved and its duration (more than 75 °C for 20 days), something never observed before in our tests. We concluded that the higher the moisture concentration, the higher the maximum temperature reached and the longer the permanence at high temperature. Following these indications, we continued monitoring the sludge with 29.5% initial moisture, by adding further water, given
135
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
T external
T up
T center
80
Addition of water
Addition of water
70
Temperature (°C)
60 50 40 30 20 10 0
0
10
20
30
40
50
60
70
80
90
100
Time (d) Fig. 8. Temperature evolution after subsequent addition of water to ordinary sludge (89% SS) after it completed a complete self-heating cycle and cooled down to ambient temperature.
T sludge
T external
250 self -combustion and damage at TC
Temperature (°C)
200
150
100
50
0 0
5
10
15
20
25
30
35
40
Time (d) Fig. 9. Self-heating of dried sludge produced in a transient drying operation.
that more than 20 days above 70 °C causes a significant evaporation. Water addition to the sludge in the bag is not really simple, nor so effective, given that the solids are rather hydrophobic, because of some heavy hydrocarbon and silicons in it, and the granules compact in the bag, resulting in a poorly water permeable texture. Nevertheless, 40 L of water was poured on top of the solids in the bag, supporting its distribution by forming a concave upper surface. Several additions were made, with the same amount of water. Results are shown in Fig. 7. Interestingly, at any water addition a further re-activation occurs, within a few days from the addition, leading to significant increases of temperature. Notwithstanding some discontinuities in
the data acquisition, it is clear that reactivation is not progressively vanishing and it occurs also after quite a long time (4 months). Such a dramatic role of water, in air permeable bags, may have different explanations. As already discussed, its role in supporting biological activity, though likely, due to the solids nature, would be surprising considered the time scale of activation and the temperatures that can be achieved. More likely, water’s role is connected to exothermic reactions, where it can be a reagent (such as hydrations) or oxygen carrier. At the same time, water evaporation should act as a cooling process, contrasting re-heating. To further investigate the role of water, we carried out a test on a standard (11% moisture) sludge, monitoring the temperature in
136
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
Fig. 10. Pictures of the sludge after loss of containment due to exceptional self-heating (visible image, left; infrared image of the framed region of the visible image, right).
different positions (see Fig. 2). We concentrated on the temperature evolution after subsequent water additions made after the solid already underwent the first phase of self-heating and cooled down to ambient temperature. Fig. 8, shows the result of this test, spanning over more than 3 months. We reported only the temperature at the center and midway between the center and the upper surface (approx. 40 cm from top) for sake of clarity. Ambient temperature (T external) is also reported, indicating both the daily and the seasonal excursion. Reactivation is evident also on this ordinary sludge, and apparently with increasing intensity (max. temperatures keep growing after each subsequent addition of water). The temperature above the center is always higher than in the center, being that water comes from the top. We cannot see any correlation with environmental temperature of the center temperature, slowly varying, and questionable correlation with the more external measurements, where more fluctuations are reported indeed. However, oscillations in the upper part of the bags occur at a frequency uncorrelated with either the daily one of that of the period (colder or warmers weeks), suggesting a different origin. Due to the unpredictable texture of the solids in the bag, with cracks together with tightly packed zones, we speculate that the penetration of water (and oxygen) progresses irregularly and unevenly. As expected, the thermal inertia in the center is larger than in the upper part; both heating and cooling are slower, suggesting the dual role of solid, in insulating the inner parts and dissipating heat with the environment, by both convection at the outer surface and evaporation also from inner zones.
3.3. Role of unsteady drying stages A breakthrough in the experiments occurred after an unexpected event. A forced stop in the drying operation of lines L3 and L4 caused some solids to remain within the shutting-down drier, for a time longer than normal. The resulting material, when packed in some big–bags (without internal gas-tight PE liner), was set aside for observation. Early, odor emissions from these bags was quite unusual, its outer temperature exceptionally high, both effects constantly increasing until the bag collapsed after one week. Spot temperature measurements identified local values in excess of 250 °C. Gas analyses with portable instruments detected high concentrations of carbon monoxide. It was clear that the drying stage, while not significant in determining the self-heating behavior at normal drying conditions (see Fig. 5), can cause a dramatic reactivity increase in the products,
under unconventional operations. These can be characterized by a longer residence time in contact with a gas at low O2 content, then likely to abound in reducing gases (e.g. CO) and longer contact with high local temperature (close to heat exchanging surfaces). We identified this circumstances as transient operative conditions. At a later occurrence of such transient conditions, we monitored the temperature in the sludge put in a gas-permeable big–bag. Temperature in the center during approx. one month is shown in Fig. 9. The initial heating compares well with ordinary sludge, but then temperature keeps increasing, at a progressively faster rate up to an already impressive value of approx. 160 °C. Then heating continues, but at a much smaller pace, suggesting a changing in mechanism, either chemical (a different set of reactions) or physical (heat or mass transport rate). Above 200 °C, a point is reached where the bag suddenly collapsed. The bunker of solids after loss of containment is shown in Fig. 10, both in visible and infrared spectra (thermal camera, Fluke, Ti 35). The color of the sludge changed significantly, from the original black to brown in large portions. Brownish area are also associated to the higher temperatures, according to the semiqualitative indication of the IR picture, up to an estimate of 450 °C in some surface areas. Interestingly, activity of even the most reactive solids obtained under transient operations can be prevented or completely controlled by stopping any contact of air with the solids, as actually verified by completely insulating with a thermoretractable film big–bags, otherwise able to heat up until combustion. Removal of the film resumes the original reactivity. Though evident that self-heating, even in its milder evolution, is caused by chemical reactions, possibly, but very marginally supported by biological activity, the exact mechanism deserves a systematic investigation, well beyond the simple monitoring of temperature. This will be illustrated in a separate report. At the same time, the potential hazard of transient operations of the driers was clearly spotted. So far, the best management of these circumstances is quite simple and effective. Products of unsteady drying stages are separately collected and poured upstream in the treatment process, in the industrial influent before the screening and grit removal.
4. Conclusions We studied the occurrence of spontaneous self-heating of sludge after drying by means of temperature measurements in single big–bags of approx. 1 t capacity. The tests allowed a better
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
understanding of the dynamics and factors influencing the selfheating phenomenon observed in the dedicated landfill. Sludge is rather peculiar because of its origin, from primary and biological treatments of both municipal and industrial wastewater, the latter largely dominant, and mainly from local tanneries. Every test showed the presence of air to be a key factor for triggering and supporting the development of the exothermic reactions, suggesting that oxidations are taking place. Precluding the contact with air prevents any temperature rise. Also water is an important factor able to initiate and support self-heating, but its role is not completely clear, and possibly not unique. While expected to support biological activity, due to the solids nature, we minimize this aspect, considering the time scale of sludge activation and the temperatures that can be achieved. More likely, water is involved in exothermic reactions, where it can be a reagent (such as hydrations) or oxygen carrier. At the same time, water evaporation is expected to be a major temperature regulating process. In any case, water can maintain the exothermic behavior for longer time and new additions cause reactivation of sludge. However, differently from air, water is not required for self-heating to manifest. The drying condition and technology does not seem to affect the dried sludge reactivity insofar dryers operate normally. All the material tested could achieve a maximum temperature between 70 and 90 °C before cooling spontaneously. However, unusual drying operation was found to affect dramatically the self-heating activity, up to spontaneous combustion. Conditions include cycles of heating and cooling, variable oxygen concentration in contact with solids, reducing atmosphere, and high local temperature all that due to discontinuous operations at plant start-up or shut-down. While run-aways in self-heating of dry sludge from transient process could be controlled by preventing contact with air, an effective management policy was simply isolating the sludge obtained at transient dryer operations, diverting from the landfill and diluting it upstream in the wastewater treatment.
137
Temperature values as well as heating time scales suggest that the nature of the exothermic process has to be mainly chemical and physical, while microbiological activity is a marginal co-factor. What is clear at this point of our investigation program gives a few suggestions to prevent or mitigate the phenomena: (i) big–bags with additional lining to limit air and water penetration; (ii) pelletizing the sludge, thus reducing the activated surface and porosity; and (iii) separate the products of unsteady drying, and keep them under observation. References Escudey, M., Arias, A., Forster, J., Moraga, N., Zambra, C., Chang, A.C., 2008. Sewage sludge self-heating and spontaneous combustion. Field, Lab. Numer. Studies, High Temp. Mater. Process. 27–5, 339–346. Fu, Z.M., Li, X.R., Koseki, H., 2005. Heat generation of refuse derived fuel with water. J. Loss Prevent. Process Ind. 18, 27–33. Gholamifard, S., Eymard, R., 2009. An Empirical Relationship Between Waste Water Content, Density and Biogas Production in Reconstituted Municipal Solid Waste: a Laboratory Scale Experiment. Veolia Environnement, France. Li, X.R., Lim, W.S., Iwata, Y., Koseki, H., 2008. Thermal behavior of sawage sludge derived fuels. Therm. Sci. 12, 137–148. Moqbel, S., Reinhart, D., Chen, R.H., 2010. Factors influencing spontaneous combustion of solid waste. Waste Manage. 30, 1600–1607. Poffet, M., Kaeser, K., Jenni, T.A., 2008. Thermal runaway of dried sewage sludge granules in storage tanks. Chimica 62, 29–34. Riley, J.T., Reasoner, J.W., Fatemi, S.M., Yates, G.S., 1987. Self-Heating of Coal Barges, Conference. 193. National meeting of the American Chemical Society, Denver, USA, 32–1. Shea, F.L., Hsu, H.L., 1972. Self-heating of carbonaceous material. Ind. Eng. Chem. Prod. Res. Dev. 11–2, 184–187. Walker, I.K., 1967. The role of water in spontaneous combustion of solid. Fire Res. Abstr. Rev. 9, 5–22. Yasuhara, A., Amano, Y., Shibamoto, T., 2010. Investigation of the self heating and spontaneous ignition of refuse derived fuel (RDF) during storage. Waste Manage. (Oxford) 30, 1161–1164. Zhao, X., Jiang, J., Meng, Y., 2006. Mechanism and influencing Factors of spontaneous combustion of oil tank containing sulfur. In: International Symposium on Safety Science and Technology, Changsha, China, pp. 1399– 1403.
Contents lists available at SciVerse ScienceDirect
Waste Management journal homepage: www.elsevier.com/locate/wasman
Self-heating of dried wastewater sludge M. Zerlottin a, D. Refosco a, M. Della Zassa b, A. Biasin b, P. Canu b,⇑ a b
Acque del Chiampo, SpA, Via Ferraretta, 20, 36071 Arzignano (VI), Italy Dept. of Industrial Engineering, University of Padua, Via Gradenigo, 5/a, 35131 Padova, Italy
a r t i c l e
i n f o
Article history: Received 27 February 2012 Accepted 22 August 2012 Available online 6 October 2012 Keywords: Self-heating Self-combustion Thermal runaway Wetting heat Tannery sludge
a b s t r a c t We experimentally studied the occurrence of spontaneous self-heating of sludge after drying, to understand its nature, course and remediation. The sludge originates from primary and biological treatment of both municipal and industrial wastewater, the latter largely dominant (approx. 90% total organic carbon, mainly from local tanneries). Dried sludge is collected into big–bags (approx. 1.5 m3) and landfilled in a dedicated site. After several years of regular operation of the landfill, without any management or environmental issue, indications of local warming emerged, together with smoke and smelling emissions, and local subsidence. During a two year monitoring activity, temperatures locally as high as 80 °C have been detected, 6–10 m deep. Experiments were carried out on large quantities of dried sludge (1 t), monitoring the temperature of the samples over long periods of time (months), aiming to reproduce the spontaneous self-heating, under different conditions, to spot enhancing and damping factors. Results demonstrate that air is a key factor to trigger and modulate the self-heating. Water, in addition to air, supports and emphasizes the heating. Unusual drying operation was found to affect dramatically the self-heating activity, up to spontaneous combustion, while ordinary drying conditions yield a sludge with a moderate self-heating inclination. Temperature values as well as heating time scales suggest that the exothermic process nature is mainly chemical and physical, while microbiological activity might be a co-factor. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Wastewater treatment plants yield a large quantity of sludge in addition to purified water. Disposing of sludge may be a significant issue. Dehydrating and drying processes allow to largely reduce the sludge mass, before landfilling. Questions arise about the chemical and biological stability of dried sludge. In the plant of our interest, quite large indeed, dried sludge is collected into PP, with waterproof PE internal layer, big–bags with a volume of approx. 1.5 m3. Bags are subsequently landfilled in a dedicated site. After several years of regular operation of the landfill, without any management or environmental difficulty, local temperature unusually high have been measured. At the same time, landfill produced smoke and malicious odor. Locally, quite significant subsidence was measured, up to 2 m. After a systematic monitoring campaign, lasting more than two years, temperatures locally as high as 80 °C have been detected, in specific zones, typically 6–10 m deep. Local surface subsidence up to 2 m was measured, over a total depth of approx. 12–14 m. While surprised by the significance of the process going on in the landfill, we started a systematic investigation, beginning with a literature survey on comparable scenarios. The self-heating on ⇑ Corresponding author. Tel./fax: +39 0498275463. E-mail address: [email protected] (P. Canu). 0956-053X/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.wasman.2012.08.014
a large scale of different materials (carbonaceous material, MSW, municipal sewage sludge) has been reported (Riley et al., 1987; Moqbel et al., 2010; Escudey et al., 2008). Riley et al. (1987) focused on coal in barges, clearly admitting that exhotermic oxidation reactions take place, stressing the role of air. Moqbel et al. (2010) revert to laboratory to identify the auto-ignition temperature of synthetic MWS; they already observed the key role of oxygen to support chemical oxidation reactions. Escudey et al. (2008) carried out field tests on sewage sludge piles extending several meters, for 20 weeks, confirming that temperature could rise up to 90 °C, never causing self combustion, though reporting that it was observed in landfills in their region. In the case of our dry sludge, the first hypotheses concerns the occurrence of some extraordinary biological activity. Literature reports contrasting opinions about that. Li et al. (2008) explicitly support the quantitative generation of heat by biological processes. Others (Poffet et al., 2008; Gholamifard and Eymard, 2009), claim that bacterial activity requires a longer induction time (days or weeks) and sufficient moisture degree. Finally, Yasuhara et al. (2010) like Fu et al. (2005) definitely exclude any connection between self-heating and the bacterial metabolism. The residual moisture in our case is always less than 15% (typically 10%), to prevent the onset of biological processes. However, loss of containment by a big–bag in the landfill, causes sludge to rehydrate, possibly triggering some biological activity. This activity, while
130
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
initially aerobic, can evolve in anaerobic where oxygen lacks, with emission of biogas. Because of the measured temperatures, up to 80 °C, we hardly believe that biological processes are the principal ones taking place. The aerobic processes are known to operate up to a maximum temperature of 70–75 °C, while the anaerobic processes are not active above 60–65 °C. Furthermore, lot of water to rehydrate the sludge is required to provide a suitable environment for biological activity. The landfill is progressively filled by segregating limited deposition areas with LDPE sheets, to repair from rain, preventing quantitative rehydration to take place. Self-heating is likely to occur because of hydration reactions involving salts and oxides in the inorganic fraction, as discussed by Fu et al. (2005) and Walker, 1967. They measured instantaneous heating of RDF and other solids after the addition of water because of what they called ‘‘wetting heat’’. The sudden heating is not compatible with any sort of growing bacterial activity. They proved that the temperature rose from ambient only after the addition of water. However, the hydration heat, if any exists in our case, can be easily overwhelmed by sensible and latent evaporation heat in case a large amount of water is used, as we experienced in remediation attempts successfully applied to the landfill. We can exclude accidental or intentional ignition (fires, cigarettes, etc.) for two reasons. The dried sludge is not easily ignited and any ignition should develop mostly locally, while heating was observed in different zones within the landfill, with the highest temperatures in the deep of the waste. The most likely cause was suggested by Poffet et al. (2008) who conjectured the presence of pyrophoric substances, able to chemically react with oxygen and water very exothermically. Authors (Li et al., 2008) observed self-heating, sometimes leading to self-combustion, of dried sludge from municipal wastewater treatments. They measured temperatures in excess of 600 °C. We were extremely surprised both by the resemblance of the case described in Poffet et al. (2008) with our experience, notwithstanding the supposed difference in the sludge origin, and the extent of the consequences. Interestingly, they remarked a significant presence of sulfur and iron, which is similar to our sludge, though the origin differs. Moreover, awareness of the possibility of fires development in dried sludge bunkers remains apparently underestimated in the management of solid wastes. It was certainly the case at the time of our landfill design, which occurred much earlier that 2008, when (Poffet et al., 2008) was published, but still a large uncertainty remains about the actual chemistry and likelihood of fire occurrence, in practice. Poffet et al. (2008) carried out further small scale investigations in small isolated vessels (Dewar type), as already suggested by Shea and Hsu (1972) reporting that an oxygen flow causes a quicker and more intense heating of the sludge than an inert flux. They also observed that water supports the heating, already suggesting that exothermic reactions could occur, like hydration of metals transition salts. It is known indeed (Zhao, 2006) that the ferrous sulfide, FeS, largely present in the sludge, can react with oxygen (air) at ambient, temperature, with a significant heat of reaction (up to 2500 kJ/mol, according to literature). As expected, the smaller the granules size, the faster the exothermic reactions develop, because of the higher interphase (gas–solids) surface. Also the environmental temperature affects the maximum temperature that solids can achieve, apparently because of a competition between heating by reactions and cooling by heat transfer to the ambient (with lower ambient temperature, the thermal dispersion increases). Being able to reproduce the self-heating of the sludge under controlled conditions is a prerequisite to understand the chemical and physical mechanism, eventually leading to control the heat release and to prevent run-aways. Here we report about large, big– bag scale investigations highlighting the basic features and magnitude of the self-heating phenomena.
What was studied here could be considered one of different phenomena, including the so called smoldering, not well understood yet, though amenable to dramatic consequences. Definitely, they deserve a scientific investigation. Here we report the beginning of our study, i.e. observations on the large scale.
2. Materials and methods 2.1. Sludge Experiments focused on dried sludge, resulting from a single wastewater treatment plant, mainly received from the local industrial tannery district. Sludge is collected both from the primary settling tank (2/3 of total solids produced in the plant are from raw wastewater) and from floatation (due to bacterial activity). The plant of our interest is quite large (1.5 millions equiv. inhabitants) resulting in a total annual mass flow of sludge suspension, at approx. 95% water, in excess of 420,000 tons/y. It is evident the need of reducing the volume. A first dewatering process raises the solids concentration of sludge to 25–32%, then drying brings it to an average of 89% in ordinary conditions. Still, the average production rate of dried sludge approaches 72 tons/d. Details of dewatering and drying are worth discussing, because they may play a role in determining the final activity of the resulting solids. Dewatering is mechanically achieved by filterpresses, where water-laden sludge is squeezed in several tissue-confined chambers. Before filtering, ferrous chloride and organic flocculant are added to the suspension; the first one binds the sulfide ions as insoluble ferrous sulfide, to limit the emission of H2S in working areas and helps with the latter the dewatering capacity of sludge, improving the flocs aggregation thus facilitating the release of water. Dewatered sludge is conveyed to two storage silos (100 m3 each) through screw conveyors. Subsequent drying is achieved thermally, by different methods. Two equal lines, L1 and L2, each one with a 4 t/h evaporation capacity, operate with direct solids heating by convection, with hot gas produced by a natural gas burner. A low concentration of oxygen (7–12%) is maintained in the drier to reduce the dust explosion hazard. Two additional equal lines, L3 and L4, each one with a 2.6 t/h evaporation capacity, were recently added. These are based on thin film evaporation down hot surfaces, externally heated by heat transfer oil. During construction and start-up of L3 and L4 lines, a support line, L5, with a 2 t/h evaporation capacity, was in operation, and produced some of the tested sludge. This combined both direct and indirect drying methods at the same time, both realized as described above. Eventually, sludge in the form of powder with granules of approx. 2 mm mean size, with 85–92% dry solid content and 0.7 kg/ L apparent density, is stored in big–bags subsequently moved to a dedicated landfill. The average composition is shown in Table 1 for the most significant elements. Note the low concentration of heavy metals except for chromium, largely used in the tannery activities, reaching up to 4%wt of solid substance. Iron (up to 2 wt.%) is due to additions at different treatment stages: as FeCl3 after biological treatment, to remove phosphorous from wastewater and improve clarification, furthermore it is added as FeCl2 before dewatering, as discussed above. Another important figure is the total sulfur content, up to 2 wt.%; compared with the small concentration of sulfate and sulfide, it suggests the presence of elemental sulfur. Sludge has a peculiar odor, changing from dewatered to dried; its further change upon self-heating indicates the occurrence of chemical transformations, to a different extent. Due to its composition and origin, mainly industrial (tannery) wastewaters, sludge must be disposed according to specific
M. Zerlottin et al. / Waste Management 33 (2013) 129–137 Table 1 Average composition of dried sludge, based on solid substance (SS). Analyses
Units
pH Dry solids content (@105 °C) Volatile solids content (@600 °C)
– % %
Total oily substances Total nitrogen (TKN) Total chromium Total sulfur Iron Total hydrocarbon Phosphorus Aluminum Sulfate Silicon Zinc Sulfide Manganese Copper Nickel Arsenic Lead Cadmium
g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg g/kg
Quantity 8.5 88.5 74.9
(SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS) (SS)
132.82 57.23 35.64 19.89 16.48 13.61 9.71 3.01 2.49 2.03 1.39 0.29 0.19 0.05 0.03 <0.023 0.01 <0.001
hazardous waste management regulations. At the present, a landfill totally dedicated to this waste is used, while technical studies to further reduce its volume, exploiting its energetic value (12– 16 MJ/kg) are under way. Over the decades, the company realized and managed nine landfills of its own property, some still receiving sludge while others reached their maximum capacity. At this point, it is covered with different layers: gravel, clay, HDPE sheets, soil and topsoil to prevent infiltration in the waste mass and kept under control for 30 years or more. At the present, dried sludge is disposed of in a single, dedicated landfill (D9), with a capacity of 366,000 m3 over a depth of 12 m. D9 has a waterproof clay shell, covered by a HDPE geomembrane. A network of perforated pipes collects the rain water to drainage wells. Big–bags are orderly placed in the landfill, as shown in Fig. 1; islands of approx. 600 bags are further segregated by means of LDPE sheets, to protect from UV and other atmospheric agents. Big–bags are made
131
of a PP tissue, with treatments against UV; consequently, they are gas permeable. They are then provided with an additional gas-tight PE internal layer (liner). 2.2. Self-heating experiments Experiments to reproduce and characterize the self-heating process were carried out at the big–bag scale, thus using large quantities of dried sludge (1 t). Most tests are carried out on a single big–bag of sludge, isolated inside an industrial enclosure where environmental conditions are more stable. A few test involved more than one bag, closely packed, to investigate cooperative effects. Measurements involve monitoring the temperature over long periods of time (up to several months). Either a single temperature, approximately in the center of the sludge mass, or several points are monitored, Fig. 2. Occasionally, semiquantitative gas emission measurements and IR images were collected. The sludge is poured into the big–bags after drying. The sludge in each monitored bag was separately analyzed, seeking correlations between different reactivity and composition. We investigated several factors: the drying technology, effects of gas permeation through the bag, aeration before filling, water addition, heat dissipation, and environmental influences. Overall, we monitored the temperature in 75 single big–bags, for a total (cumulated) of 90 months of observation, resulting in very extensive experimental campaign during a period of more that 16 months. In the following we illustrate the most significant results. 3. Results and discussion 3.1. Role of air The first issue that was investigated is the role of the inner PE liner gas-tight. Intuition suggested that it could be a severe limitation for water penetration in case excessively hot solids have to be cooled. The same intuition suggests that the PE inner bag could actually support the heating process, limiting the heat dissipation.
Fig. 1. Overview of the landfill dedicated to dried sludge (D9).
132
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
0.4m 0.75m 1.15m 1.25m
0.25m 0.5m 1m
Fig. 2. Schematic of a single big–bag with thermocouples positions.
Fig. 3. Increase of temperature in two big–bags, with and without internal liner.
Two bags, with and without inner liner, were filled with the same dry sludge, placed in a closed space and the central temperature was monitored. Results are shown in Fig. 3. Up to the 7th day, bags where kept closed, as usual in the landfill. The temperature at the end of filling was approx. 50 °C for both samples; in both cases we observed a sudden initial rise of temperature of approx. 10 °C. Surprisingly, the behavior afterwards shows that the inner PE layer kept the temperature rise fairly limited, below approx. 63 °C, while the temperature of the sludge contained in the knitted PP bag rose significantly in a few days, approaching 90 °C. Maximum temperatures are also achieved after quite different time delays, 3–4 days without internal layer, and approx. 1 day with the gas-tight liner. After peaking, temperature gradually decreases, again at a different rate. The internal liner causes a slower cooling. Complete cooling down to ambient temperature took almost 3 weeks for the single big–bag with the simple PP envelope.
The different behavior suggests that contact with air (filling the bag by pouring, having a gas permeable bag) could result in enhancement of the reactivity. To further confirm this tentative conclusion, the bag with PE internal liner was opened after clear, irreversible cooling took place, i.e. at the 7th day. The bag was opened at the top and several small holes drilled on the bag sides to allow permeation of air as much as possible, approaching the big–bag without liner. As clear form Fig. 3, reactivity soon increased enough to temporally overcome the heat dissipation, causing the declining temperature to rise again, up to a secondary maximum. Following the intuition provided by the above experiments, we investigated the effect of a prolonged contact with air before filling the bags. We carried out three tests in parallel, on the same sludge. In one case the standard, 15 min. filling time was applied. The other test achieve a longer filling time, by
133
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
Test a 15min
Test b 35min
Test c 60min
90
Temperature°C
80
70
60
50
40
30
20
0
1
2
3
4
5
6
7
8
Time (d) Fig. 4. Effect of pre-aeration before packaging the sludge into big–bags without inner liner.
100
Temperature°C
90
80
70
60
50
40
30 0
1
2
3
4
5
Time (d) Direct
Mix
Indirect
Fig. 5. Comparison between different shelf-heating in big–bags without inner liner of sludge dried with different technologies.
introducing several pauses, given that the discharge flow rate cannot be modified. Results are shown in Fig. 4. The shape of the profiles are almost identical, but the extent of pre-aeration affects the initial temperature. Moreover, we observed the double, diverging role of air causing both heat dissipation and self-heating enhancement. Results of Fig. 4 suggest that a more thorough exposure to air during filling better dissipate the residual heat content in the sludge, but it is not sufficient to modify the subsequent self heating behavior, which evolve over a much longer time scale and is determined by the capacity to aerate the sludge during that interval (i.e. having a gas permeable or tight containment, as shown in Fig. 3). Thermal behavior of sludge after these additional tests is clearly characterized by its contact with air.
With the availability of different drying technologies, such as direct (L1, L2), indirect (L3, L4) and mixed (L5) heat exchange, expected to expose the sludge to different temperature histories, we planned a comparative test, in parallel on the same dehydrated sludge. Results are shown in Fig. 5. Self-heating of sludge, when in the final bag, is not significantly depending on the drying technology, being direct or indirect heat transfer. A different behavior is observed with sludge from L5, that uses a mixture of both direct and indirect heating (curve labeled Mix in the figure); a lowering of the initial temperature is observed. Because of its structure, L5 discharges the dried sludge through a screw conveyor, water cooled, with external air draft; that allows the sludge to cool and maybe partially react before pouring it into the bag. Eventually, we concluded that the drying method does not affect
134
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
moisture conc. 13,3%
moisture conc. 29,5%
95 85
Temperature (°C)
75 65 55 45 35 25 15 0
5
10
15
20
25
30
35
40
45
Time (d) Fig. 6. Effect of residual moisture in the dried sludge.
95
85
Temperature (°C)
75
addition of water
65
55
45
35
25
15
Time (d) Fig. 7. Reactivation of the self-heating by subsequent additions of water.
significantly conditions.
the
self-heating
behavior,
in
ordinary
drying
3.2. Role of water Furthermore, based on intuition, we expected the self-heating to be stronger with drier solids or, seen on the other side, a larger amount of moisture to mitigate the reactivity. We check this hypothesis comparing the behavior of the same dehydrated sludge that was dried to a different extent, in the same unit (L5). Both were put in gas permeable (no internal PE liner) big–bags. Fig. 6 shows the results, clearly contradicting the presumption. While residual moisture dramatically affects the temperature increase
and intensity of self-heating, the direction was the opposite than expected. If the sludge with a residual moisture slightly above the ordinary (i.e. 13.3% instead of the average of 11%) compares well with previous results, the sludge with a remarkably larger residual moisture (29.5%) shows a much stronger self-heating effect, both in terms of maximum temperature achieved and its duration (more than 75 °C for 20 days), something never observed before in our tests. We concluded that the higher the moisture concentration, the higher the maximum temperature reached and the longer the permanence at high temperature. Following these indications, we continued monitoring the sludge with 29.5% initial moisture, by adding further water, given
135
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
T external
T up
T center
80
Addition of water
Addition of water
70
Temperature (°C)
60 50 40 30 20 10 0
0
10
20
30
40
50
60
70
80
90
100
Time (d) Fig. 8. Temperature evolution after subsequent addition of water to ordinary sludge (89% SS) after it completed a complete self-heating cycle and cooled down to ambient temperature.
T sludge
T external
250 self -combustion and damage at TC
Temperature (°C)
200
150
100
50
0 0
5
10
15
20
25
30
35
40
Time (d) Fig. 9. Self-heating of dried sludge produced in a transient drying operation.
that more than 20 days above 70 °C causes a significant evaporation. Water addition to the sludge in the bag is not really simple, nor so effective, given that the solids are rather hydrophobic, because of some heavy hydrocarbon and silicons in it, and the granules compact in the bag, resulting in a poorly water permeable texture. Nevertheless, 40 L of water was poured on top of the solids in the bag, supporting its distribution by forming a concave upper surface. Several additions were made, with the same amount of water. Results are shown in Fig. 7. Interestingly, at any water addition a further re-activation occurs, within a few days from the addition, leading to significant increases of temperature. Notwithstanding some discontinuities in
the data acquisition, it is clear that reactivation is not progressively vanishing and it occurs also after quite a long time (4 months). Such a dramatic role of water, in air permeable bags, may have different explanations. As already discussed, its role in supporting biological activity, though likely, due to the solids nature, would be surprising considered the time scale of activation and the temperatures that can be achieved. More likely, water’s role is connected to exothermic reactions, where it can be a reagent (such as hydrations) or oxygen carrier. At the same time, water evaporation should act as a cooling process, contrasting re-heating. To further investigate the role of water, we carried out a test on a standard (11% moisture) sludge, monitoring the temperature in
136
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
Fig. 10. Pictures of the sludge after loss of containment due to exceptional self-heating (visible image, left; infrared image of the framed region of the visible image, right).
different positions (see Fig. 2). We concentrated on the temperature evolution after subsequent water additions made after the solid already underwent the first phase of self-heating and cooled down to ambient temperature. Fig. 8, shows the result of this test, spanning over more than 3 months. We reported only the temperature at the center and midway between the center and the upper surface (approx. 40 cm from top) for sake of clarity. Ambient temperature (T external) is also reported, indicating both the daily and the seasonal excursion. Reactivation is evident also on this ordinary sludge, and apparently with increasing intensity (max. temperatures keep growing after each subsequent addition of water). The temperature above the center is always higher than in the center, being that water comes from the top. We cannot see any correlation with environmental temperature of the center temperature, slowly varying, and questionable correlation with the more external measurements, where more fluctuations are reported indeed. However, oscillations in the upper part of the bags occur at a frequency uncorrelated with either the daily one of that of the period (colder or warmers weeks), suggesting a different origin. Due to the unpredictable texture of the solids in the bag, with cracks together with tightly packed zones, we speculate that the penetration of water (and oxygen) progresses irregularly and unevenly. As expected, the thermal inertia in the center is larger than in the upper part; both heating and cooling are slower, suggesting the dual role of solid, in insulating the inner parts and dissipating heat with the environment, by both convection at the outer surface and evaporation also from inner zones.
3.3. Role of unsteady drying stages A breakthrough in the experiments occurred after an unexpected event. A forced stop in the drying operation of lines L3 and L4 caused some solids to remain within the shutting-down drier, for a time longer than normal. The resulting material, when packed in some big–bags (without internal gas-tight PE liner), was set aside for observation. Early, odor emissions from these bags was quite unusual, its outer temperature exceptionally high, both effects constantly increasing until the bag collapsed after one week. Spot temperature measurements identified local values in excess of 250 °C. Gas analyses with portable instruments detected high concentrations of carbon monoxide. It was clear that the drying stage, while not significant in determining the self-heating behavior at normal drying conditions (see Fig. 5), can cause a dramatic reactivity increase in the products,
under unconventional operations. These can be characterized by a longer residence time in contact with a gas at low O2 content, then likely to abound in reducing gases (e.g. CO) and longer contact with high local temperature (close to heat exchanging surfaces). We identified this circumstances as transient operative conditions. At a later occurrence of such transient conditions, we monitored the temperature in the sludge put in a gas-permeable big–bag. Temperature in the center during approx. one month is shown in Fig. 9. The initial heating compares well with ordinary sludge, but then temperature keeps increasing, at a progressively faster rate up to an already impressive value of approx. 160 °C. Then heating continues, but at a much smaller pace, suggesting a changing in mechanism, either chemical (a different set of reactions) or physical (heat or mass transport rate). Above 200 °C, a point is reached where the bag suddenly collapsed. The bunker of solids after loss of containment is shown in Fig. 10, both in visible and infrared spectra (thermal camera, Fluke, Ti 35). The color of the sludge changed significantly, from the original black to brown in large portions. Brownish area are also associated to the higher temperatures, according to the semiqualitative indication of the IR picture, up to an estimate of 450 °C in some surface areas. Interestingly, activity of even the most reactive solids obtained under transient operations can be prevented or completely controlled by stopping any contact of air with the solids, as actually verified by completely insulating with a thermoretractable film big–bags, otherwise able to heat up until combustion. Removal of the film resumes the original reactivity. Though evident that self-heating, even in its milder evolution, is caused by chemical reactions, possibly, but very marginally supported by biological activity, the exact mechanism deserves a systematic investigation, well beyond the simple monitoring of temperature. This will be illustrated in a separate report. At the same time, the potential hazard of transient operations of the driers was clearly spotted. So far, the best management of these circumstances is quite simple and effective. Products of unsteady drying stages are separately collected and poured upstream in the treatment process, in the industrial influent before the screening and grit removal.
4. Conclusions We studied the occurrence of spontaneous self-heating of sludge after drying by means of temperature measurements in single big–bags of approx. 1 t capacity. The tests allowed a better
M. Zerlottin et al. / Waste Management 33 (2013) 129–137
understanding of the dynamics and factors influencing the selfheating phenomenon observed in the dedicated landfill. Sludge is rather peculiar because of its origin, from primary and biological treatments of both municipal and industrial wastewater, the latter largely dominant, and mainly from local tanneries. Every test showed the presence of air to be a key factor for triggering and supporting the development of the exothermic reactions, suggesting that oxidations are taking place. Precluding the contact with air prevents any temperature rise. Also water is an important factor able to initiate and support self-heating, but its role is not completely clear, and possibly not unique. While expected to support biological activity, due to the solids nature, we minimize this aspect, considering the time scale of sludge activation and the temperatures that can be achieved. More likely, water is involved in exothermic reactions, where it can be a reagent (such as hydrations) or oxygen carrier. At the same time, water evaporation is expected to be a major temperature regulating process. In any case, water can maintain the exothermic behavior for longer time and new additions cause reactivation of sludge. However, differently from air, water is not required for self-heating to manifest. The drying condition and technology does not seem to affect the dried sludge reactivity insofar dryers operate normally. All the material tested could achieve a maximum temperature between 70 and 90 °C before cooling spontaneously. However, unusual drying operation was found to affect dramatically the self-heating activity, up to spontaneous combustion. Conditions include cycles of heating and cooling, variable oxygen concentration in contact with solids, reducing atmosphere, and high local temperature all that due to discontinuous operations at plant start-up or shut-down. While run-aways in self-heating of dry sludge from transient process could be controlled by preventing contact with air, an effective management policy was simply isolating the sludge obtained at transient dryer operations, diverting from the landfill and diluting it upstream in the wastewater treatment.
137
Temperature values as well as heating time scales suggest that the nature of the exothermic process has to be mainly chemical and physical, while microbiological activity is a marginal co-factor. What is clear at this point of our investigation program gives a few suggestions to prevent or mitigate the phenomena: (i) big–bags with additional lining to limit air and water penetration; (ii) pelletizing the sludge, thus reducing the activated surface and porosity; and (iii) separate the products of unsteady drying, and keep them under observation. References Escudey, M., Arias, A., Forster, J., Moraga, N., Zambra, C., Chang, A.C., 2008. Sewage sludge self-heating and spontaneous combustion. Field, Lab. Numer. Studies, High Temp. Mater. Process. 27–5, 339–346. Fu, Z.M., Li, X.R., Koseki, H., 2005. Heat generation of refuse derived fuel with water. J. Loss Prevent. Process Ind. 18, 27–33. Gholamifard, S., Eymard, R., 2009. An Empirical Relationship Between Waste Water Content, Density and Biogas Production in Reconstituted Municipal Solid Waste: a Laboratory Scale Experiment. Veolia Environnement, France. Li, X.R., Lim, W.S., Iwata, Y., Koseki, H., 2008. Thermal behavior of sawage sludge derived fuels. Therm. Sci. 12, 137–148. Moqbel, S., Reinhart, D., Chen, R.H., 2010. Factors influencing spontaneous combustion of solid waste. Waste Manage. 30, 1600–1607. Poffet, M., Kaeser, K., Jenni, T.A., 2008. Thermal runaway of dried sewage sludge granules in storage tanks. Chimica 62, 29–34. Riley, J.T., Reasoner, J.W., Fatemi, S.M., Yates, G.S., 1987. Self-Heating of Coal Barges, Conference. 193. National meeting of the American Chemical Society, Denver, USA, 32–1. Shea, F.L., Hsu, H.L., 1972. Self-heating of carbonaceous material. Ind. Eng. Chem. Prod. Res. Dev. 11–2, 184–187. Walker, I.K., 1967. The role of water in spontaneous combustion of solid. Fire Res. Abstr. Rev. 9, 5–22. Yasuhara, A., Amano, Y., Shibamoto, T., 2010. Investigation of the self heating and spontaneous ignition of refuse derived fuel (RDF) during storage. Waste Manage. (Oxford) 30, 1161–1164. Zhao, X., Jiang, J., Meng, Y., 2006. Mechanism and influencing Factors of spontaneous combustion of oil tank containing sulfur. In: International Symposium on Safety Science and Technology, Changsha, China, pp. 1399– 1403.