Pacific Northwest National Laboratory, Richland, WA 99352, USA
Robert S. Weber
Show less
History+
Received
Accepted
Published Online
2017-11-14
2018-05-17
2018-05-24
PDF
(1275KB)
Abstract
We have adapted and characterized electrolysis reactors to complement the conversion of regional- and community-scale quantities of waste into fuel or chemicals. The overall process must be able to contend with a wide range of feedstocks, must be inherently safe, and should not rely on external facilities for co-reactants or heat rejection and supply. Our current approach is based on the upgrading of bio-oil produced by the hydrothermal liquefaction (HTL) of carbon-containing waste feedstocks. HTL can convert a variety of feedstocks into a bio-oil that requires much less upgrading than the products of other ways of deconstructing biomass. We are now investigating the use of electrochemical processes for the further conversions needed to transform the bio-oil from HTL into fuel or higher value chemicals. We, and others, have shown that electrochemical reduction can offer adequate reaction rates and at least some of the necessary generality. In addition, an electrochemical reactor necessarily both oxidizes (removes electrons) on one side of the reactor and reduces (adds electrons) on the other side. Therefore, the two types of reactions could, in principle, be coupled to upgrade the bio-oil and simultaneously polish the water that is employed as a reactant and a carrier in the upstream HTL. Here, we overview a notional process, the possible conversion chemistry, and the economics of an HTL-electrochemical process.
Robert S. Weber, Johnathan E. Holladay.
Modularized Production of Value-Added Products and Fuels from Distributed Waste Carbon-Rich Feedstocks.
Engineering, 2018, 4(3): 330-335 DOI:10.1016/j.eng.2018.05.012
If waste, exemplified here as urban waste, could be converted into a liquid fuel or a chemical product, then, in principle and in aggregate, it could displace a noticeable fraction of the petroleum now consumed in the world (Table 1) [1–3]. More significantly, it would transform a costly problem of disposal (from less than 20 to over 200 USD·t−1, depending on population density [4]) into a positive economic resource. The values in the table for the now-wasted resources represent their total enthalpy of combustion; their conversion to fuel would yield a significantly smaller (∼50%) energy stream.
The recoverable heating value would likely be less than the total heating value owing to parasitic process demands and product selection. The value cited above represents complete conversion as an upper limit, as has been used in estimates by others [5].
However, three major techno-economic considerations impede the conversion of waste into a more valuable product: the heterogeneity of the material, its geographic dispersion, and its local scale.
For example, urban waste consists of a heterogeneous mixture of materials, only some of which might be converted into carbon-containing fuels or chemicals [6]. The amount and composition of urban waste varies significantly with local income and region [2].
The material is widely dispersed and is generated in each locale at a rate that is small compared with the scale of petroleum refining and chemical manufacturing (Table 2) [2,3,5,7–12]. For reference, consider that there are only about 500 petroleum refineries in the world [7], whereas the number of conventional waste-treatment facilities is likely comparable to the number of municipalities. The latter quantity is difficult to estimate because the term “municipality” is not used consistently across the world. As a surrogate, consider that in this set of upper income countries—Australia, Canada, France, the United Kingdom, and the United States—there are municipal wastewater-treatment plants serving populations that range in size from 4000 to about 50 000 people. For example, in the United States, the population of 325 million is served by about 15 000 wastewater-treatment plants, so the average plant serves a community of about 22 000 people. Plants serving populations of a million people exist. Here, we seek to estimate the number of facilities that might host waste-to-energy convertors. If the lowest density of the wastewater-treatment plants (1 plant per 50 000 people) is applied to other waste depots, then there would be more than 100 000 sites around the world (7.3 billion people/50 000 people served per site).
Moreover, as shown in Table 2, the amount of material that might be converted at each site is large enough to be bothersome (i.e., tons to kilotons per day) but small enough to require technology that will be unlike that used in the industrial refining of petroleum or the manufacture of chemicals.
An engineering solution to the valorization of waste will require a robust technology that can be practiced economically and, of course, in a safe and environmentally friendly manner, in remote locations near to the source of the feedstock. Those constraints could be satisfied with operating conditions close to ambient temperature and pressure, implying that intensified unit operations, particularly those that do not rely merely on thermal activation, would be preferred. The large number of possible sites motivates the consideration of a modular, manufacturable design rather than a site-specific, stick-built architecture characteristic of traditional refining and chemical plants.
2. Notional process design
The core of a waste-to-fuel or -chemical facility would include an omnivorous front end, upgrading stage(s), and separations. Experience at Pacific Northwest National Laboratory (PNNL) suggests that hydrothermal liquefaction (HTL) can transform a wide range of carbon/hydrogen/oxygen (C/H/O)-containing streams (e.g., algae [13], sewage sludge [14], or wood pulp [15]) into a bio-oil that is more thermally stable and has significantly higher heating value than bio-oil produced by pyrolysis [15–17]. We have demonstrated [18] a modular, pilot-scale system whose throughput corresponds roughly to 10% of the lower range of throughputs presented in Table 2, as shown in Fig. 1. The photograph shows, from left to right, modules for preparing and handling the feed, HTL (beneath the yellow hoods), and recovery of the bio-oil. This system can produce about 16 L·h−1 (≈ 2.4 bbl·d−1) from organic sludge.
Because the hydrothermally liquefied intermediate is dispersed in an electrically conductive polar medium (water, alcohols plus dissolved salts), it is natural to consider electrochemically activated processes to hydrogenate the bio-oil and to electrolytically condense the plentiful carboxylic acids produced by the hydrolysis of the carbon–carbon (C—C) bonds in the feed. Hydrogenations of the components of bio-oil (Fig. 2) have been demonstrated [19–22]. Electrochemical condensation reactions are well known from classical organic chemistry [23].
The separation of nonpolar products (e.g., hydrocarbons) from a polar, possibly surfactant-laden aqueous stream should be possible, but because we have not yet investigated it, it is represented only notionally on a hypothetical process flow diagram (Fig. 3). We have identified several classes of technical challenges that could impede the introduction of such technology (Table 3).
The third challenge in the table regarding water cleanup will be discussed in Section 4.
Electrochemically activated conversions enable process intensification: They can afford rates of conversion that satisfy Weisz’s window of realistic rates [24], namely on the order of 1 mol·(m3·s)−1 even at ambient conditions, because small, readily applied potentials, say 0.1 V, correspond to an energy field that would be obtained in a thermal bath at a temperature about 1000 K higher.
For reactions that are electrochemically activated:
where Re is the rate of the electrochemically activated reaction, Re,ref is the rate of a reference electrochemically activated reaction, Δη is the difference in overpotential for the two reactions, F is the Faraday constant, R is the gas constant, and T is temperature. At a low operating temperature of, say, 350 K for a two-electron process, a 0.1 V change in the applied overpotential, η, from that of a reference state raises the reaction rate by a factor of about 800.
For reactions that are thermally activated, with an apparent Arrhenius activation energy, Ea:
where Rt is the rate of the thermally activated reaction, Rt,ref is the rate of a reference thermally activated reaction, Ea is the apparent Arrhenius activation energy, and Tref and T are the temperature at which the reference and desired reaction take place, respectively.
If Ea and Tref are representative of a catalyzed hydrodeoxygenation at, say, 50 kJ·mol−1 and 300 °C [25], then the temperature required to achieve a rate constant that is 800 times greater is about 1580 K, or 1000 K higher.
Moreover, as will be discussed below, a reticulated semi-structured electrode, such as a carbon felt, performs as a static mixer, effecting high rates of mass transfer even in creeping flow.
3. Extrapolating to flow electrochemistry
Although much of the past and current research in electrochemistry uses batch reactors, these would be cumbersome to employ even at the throughputs corresponding to the lowest entries presented in Table 2. Therefore, in designing the hypothesized process, we have assumed that the electrochemical upgrading—indeed, all the unit operations—would be continuous.
In our current work, we use flow electrolysis cells that derive from previous experience with redox flow batteries. The cells are sometimes said to have a “filter press” architecture (Fig. 4) because the anode, electrolyte membrane, cathode, flow distributors, and current collectors are stacked between thrust plates.
The electrodes, both the cathode and anode, consist of nanometer-sized metal particles supported at loadings of 5–10 wt% on a conducting graphite felt that was calcined in dilute air to make it wetted by water. In our preliminary experiments, we used an electrode devised from a direct methanol fuel cell for the anode (Ru on carbon paper), and we focused on the reduction (cathodic) reactions. The course of the reduction process depends on the substrate and the choice of metal: Aromatic aldehydes (e.g., benzaldehyde) reduce easily on Pt group metals; however, because those metals favor hydrogen atom transfer, the faradaic efficiency of the process suffers (i.e., the hydrogen that is generated can transfer to the substrate or just bubble away). We are still in the process of devising correlations for the activities of the redox metals and their interactions with a wide range of substrates (i.e., aromatic aldehydes, aliphatic aldehydes, sugars, and aromatic oxygenates such as phenol, acetophenone, and others that are found in bio-oils). Promising results have been reported from our laboratory for model compounds and for actual bio-oil [26,27].
In parallel to this work, we are characterizing the flow reactor itself to ensure that the catalyst we employ will be used effectively. For example, we have demonstrated by flow visualization in a windowed cell (Fig. 5) that even at creeping flow (Re ≪ 10), the flow profile through the reactor is plug-like.
We are in the process of measuring the effective mass transfer rates to determine whether the observed rates of conversion represent the intrinsic reaction kinetics or, rather, the rates of transport to/from the electrode surfaces. Preliminary analysis suggests that the rate of mass transfer varies with flow rate as expected [28]:
where Sh is the Sherwood number (ratio of rates of mass transfer to diffusion), Re is the Reynolds number (ratio of viscous to inertial forces), Sc is the Schmidt number (ratio of momentum and mass diffusivities), Le is the aspect ratio of the flow channel, and k ≈ 0.3. For the operating conditions we have employed, where 100% conversion corresponds to a conversion rate of 0.3 mmol·s−1, this analysis implies that the design of our reactors allows them to operate in a kinetically limited regime; that is, the rate of reaction is less than the rate of transport (Table 4) [28–30].
4. Coupling reduction and oxidation
The aqueous phase produced during HTL contains many compounds (Fig. 6) that can, in principle, be oxidized in the anode compartment of the flow electrolysis cell. Therefore, by directing the wastewater into the anode compartment and the bio-oil feed into the cathode compartment (Fig. 7), the flow electrolysis cell could simultaneously remediate the wastewater while upgrading the bio-oil. This configuration could succeed in “preventing a waste and turning it to profit,” as envisioned a century ago by one of the founders of the discipline of chemical engineering in the United States [31].
From our analysis of the hydrogen deficiency of the bio-oil and the chemical oxygen demand of the wastewater [14], we have determined that the stoichiometric ratio of oxidizable and reducible species (Table 5) could be more than sufficient to satisfy, and perhaps power, the upgrading reaction. That excess means that the electrochemical process could be net-electricity neutral and, perhaps, even an exporter of electrical power. However, the relative rates of the reactions at potentials that yield high overall conversion and selectivity have yet to be measured. Moreover, the detailed operation of the facility with respect to importing or exporting electrical power will require an optimization based on the costs of the inputs and the value of all the products, which are likely to be place and time-dependent.
5. Preliminary estimate of capital costs
We expect that the technology overviewed here could achieve significant market penetration if it could be sold at a price of around 100 000 USD·(bbl·d)−1 of capacity. That level of capital expense is low enough such that it could be paid back quickly (about three years if the products were sold at 50 USD·bbl−1 and if a credit of 50 USD·t−1 could be claimed for avoiding the waste disposal fee). The internal rate of return for that scenario would be 35% over a 10-year period. An equipment price of 100 000 USD·(bbl·d)−1 is approximately the cost of a small hydrothermal reactor unit (100 000 USD·(bbl·d)−1) plus an electrolyzer (1500 USD·(bbl·d)−1, see below) when manufactured in large quantities (m > 80):
The capital expense of a hydrothermal reactor unit, which, in essence, is a pump, a heated tube, and a hot filter, albeit constructed from corrosion-resistant alloys (Fig. 3), is about 410 000 USD·(bbl·d)−1[14] when extrapolated to 1 bbl·d−1 using a power-law extrapolation (n = 0.7) for the first plant.
We have based a preliminary estimate of the capital expense of the electrochemical upgrader on PNNL’s detailed estimate [32] for the hardware cost (Fig. 8) of a flow electrolyzer, namely, 300 USD·kW−1.
The electrolyzer would have to pass a current, I, corresponding to the stoichiometry given in Table 5:
At a cell potential of 1 V (comparable to what we typically use to perform the reductions shown in Fig. 2), the power draw would be 4.2 kW. Therefore, the cost of the electrolyzer would be about 1500 USD (300 USD·kW−1 × 4.3 kW).
A learning rate, p = 80%, consistent with mass manufacturing, lowers the total capital expenditure (CapEx) cost to about 100 000 USD·(bbl·d)−1 when the number of units, m = 80 units have been manufactured:
There would also be costs for preparation of the feed (i.e., separation, comminution) and product separation that would have to be covered in the capital budget. Still, 100 000 USD·(bbl·d)−1 appears to be a reasonable first estimate.
It should be re-emphasized that the approach envisioned here would rely on numbering up, not scaling up, the unit operations. Therefore, we expect more benefit from economies of manufacture (learning) than from economies of scale.
6. Summary
Preliminary experience and the estimates described here offer the hope that waste-to-fuel and waste-to-chemicals processes are both technically and economically feasible. The technology is being refined in current research at PNNL to develop the catalysts and the separators required to implement waste-to-fuel and waste-to-chemicals processes at the scale and in the remote locations associated with the envisaged feedstocks. We have also investigated the economics of numbering up versus the more traditional scaling up in order to address the size of the global accumulation rates of waste. That analysis will appear soon in the reports from a series of workshops sponsored by the US Department of Energy and held in 2015–2016 [33,34].
The world factbook [Internet]. Washington, DC: Central Intelligence Agency; [updated 2018 Apr 1; cited 2018 Apr 23]. Available from: https://www.cia.gov/library/publications/the-world-factbook/geos/us.html.
[2]
HoornwegD., Bhada-TataP.. What a waste: a global review of solid waste management.
[3]
MrusS.T, PrendergastC.A.. Heating value of refuse-derived fuel. In: New York: American Society of Mechanical Engineers; 1978. p. 365-370.
[4]
MortensenC.. Landfill tipping fees in California [Internet]. [cited 2017 Jul 24]. Available from: http://www.calrecycle.ca.gov/publications/Documents/1520%5C20151520.pdf
[5]
HoffmanB., NievesR.. Biofuels and bioproducts from wet and gaseous waste streams: challenges and opportunities. Jan. Report No: DOE/EE-1472 7601
[6]
LohriC.R., DienerS., ZabaletaI., MertenatA., ZurbrüggC.. Treatment technologies for urban solid biowaste to create value products: a review with focus on low- and middle-income settings. Rev Environ Sci Biotechnol. 2017; 16(1): 81-130.
[7]
Current list of crude oil refineries resources [Internet]. Global Energy Observatory; 2017 [cited 2017 Jul 18]. Available from: http://globalenergyobservatory.org/list.php?db=Resources&type=Crude_Oil_Refineries.
[8]
US Department of Agriculture. Livestock and poultry: world markets and trade.
[9]
US Department of Agriculture. Dairy: world markets and trade.
[10]
World Bank. Daily manure production per animal [Internet]. [cited 2018 Apr 23]. Available from: http://siteresources.worldbank.org/INTAPCFORUM/Resources/part3.ppt
[11]
World Bank. New data reveals uptick in global gas flaring [Internet]. [cited 2017 Jul 18]. Available from: http://www.worldbank.org/en/news/press-release/2016/12/12/new-data-reveals-uptick-in-global-gas-flaring
[12]
World Nuclear Association. Heating value of various fuels [Internet]. [cited 2017 Jul 18]. Available from: http://www.world-nuclear.org/information-library/facts-and-figures/heat-values-of-various-fuels.aspx
[13]
ElliottD.C., HartT.R., SchmidtA.J., NeuenschwanderG.G., RotnessL.J., OlarteM.V., . Process development for hydrothermal liquefaction of algae feedstocks in a continuous-flow reactor. Algal Res. 2013; 2(4): 445-454.
[14]
Snowden-SwanL.J., ZhuY., JonesS.B., ElliottD.C., SchmidtA.J., HallenR.T., . Hydrothermal liquefaction and upgrading of municipal wastewater treatment plant sludge: a preliminary techno-economic analysis. Report No: PNNL-25464 BM0102060
[15]
ElliottD.C., BillerP., RossA.B., SchmidtA.J., JonesS.B.. Hydrothermal liquefaction of biomass: developments from batch to continuous process. Bioresour Technol. 2015; 178: 147-156.
[16]
Doassans-CarrèreN., FerrasseJ.H., BoutinO., MauvielG., LédéJ.. Comparative study of biomass fast pyrolysis and direct liquefaction for bio-oils production: products yield and characterizations. Energy Fuels. 2014; 28(8): 5103-5111.
[17]
HaarlemmerG., GuizaniC., AnoutiS., DénielM., RoubaudA., ValinS.. Analysis and comparison of bio-oils obtained by hydrothermal liquefaction and fast pyrolysis of beech wood. Fuel. 2016; 174: 180-188.
Saffron CM, Li Z, Miller DJ, Jackson JE, inventors; Board of Trustees of Michigan State University, assignee. Electrocatalytic hydrogenation and hydrodeoxygenation of oxygenated and unsaturated organic compounds. United States patent US 20150008139. 2015 Jan 8.
[20]
LamC.H., LoweC.B., LiZ., LongeK., RayburnJ.T., CaldwellM.A., . Electrocatalytic upgrading of model lignin monomers with earth abundant metal electrodes. Green Chem. 2015; 17(1): 601-609.
[21]
LiZ., GaredewM., LamC.H., JacksonJ.E., MillerD.J., SaffronC.M.. Mild electrocatalytic hydrogenation and hydrodeoxygenation of bio-oil derived phenolic compounds using ruthenium supported on activate carbon cloth. Green Chem. 2012; 14(9): 2540-2549.
[22]
SinghN., SongY., GutiérrezO.Y., CamiaioniD.M., CampbellC.T., LercherJ.A.. Electrocatalytic hydrogenation of phenol over platinum and rhodium: unexpected temperature effects resolved. ACS Catal. 2016; 6(11): 7466-7470.
[23]
NollerC.R.. Chemistry of organic compounds. 3rd ed.
[24]
WeiszP.B.. The science of the possible. Chem Tech. 1982; 12: 424-425.
[25]
ZhaoH.Y., LiD., BuiP., OyamaS.T.. Hydrodeoxygenation of guaiacol as model compound for pyrolysis oil on transition metal phosphide hydroprocessing catalysts. Appl Catal A Gen. 2011; 391(1–2): 305-310.
[26]
Gutiérrez-Tinoco OY, Sanyal U, Akhade S, Lopez-Ruiz JA, Camaioni DM, Holladay J, et al. Electrocatalytic reduction of carbonyl groups in aromatic molecules: a step towards electrochemical bio-oil conversion. In: Proceedings of the 232nd ECS Meeting; 2017 Oct 1–5; National Harbor, MD, USA; 2017.
[27]
WeberR.S., HolladayJ.E., JenksC.. Modularized production of fuels and other value-added products from distributed, waste or stranded feedstocks. In: Minneapolis: American Institute of Chemical Engineers; 2017.
[28]
RalphT.R., HitchmanM.L., MillingtonJ.P., WalshF.C.. Mass transport in an electrochemical laboratory filterpress reactor and its enhancement by turbulence promoters. Electrochim Acta. 1996; 41(4): 591-603.
[29]
MikhailS.Z., KimelW.R.. Densities and viscosities of methanol-water mixtures. J Chem Eng Data. 1961; 6(4): 533-537.
[30]
DerlackiZ.J., EastealA.J., EdgeA.V.J., WoolfL.A., RoksandicZ.. Diffusion coefficients of methanol and water and the mutual diffusion coefficient in methanol-water solutions at 278 and 298 K. J Phys Chem. 1985; 89(24): 5318-5322.
[31]
LittleA.D.. Editorial. Natural resources and manufacture. Ind Eng Chem. 1909; 1(2): 61-62.
[32]
ViswanathanV., CrawfordA., StephensonD., KimS., WangW., LiB., . Cost and performance model for redox flow batteries. J Power Sources. 2014; 247: 1040-1051.
[33]
Ames Laboratory. Chemical Conversion via Modular Manufacturing: Distributed, Stranded, and Waste Feedstocks; 2015 Dec 2–4; St. Louis, MO, USA; 2015.
[34]
Ames Laboratory. Workshop on Fundamental Science Needs for Waste to Chemical Conversion. 2016 Feb 29–Mar 1; Gaithersburg, MD, USA; 2016.
PDF
(1275KB)
