Environ. Sci. Technol. 2005, 39, 5689-5693
Gaseous Mercury from Curing Concretes that Contain Fly Ash: Laboratory Measurements DANOLD W. GOLIGHTLY,† PING SUN,† CHIN-MIN CHENG,† PANUWAT TAERAKUL,† HAROLD W. WALKER,† L I N D A K . W E A V E R S , * ,† A N D DEAN M. GOLDEN‡ Civil and Environmental Engineering and Geodetic Science, The Ohio State University, 470 Hitchcock Hall, 2070 Neil Avenue, Columbus, Ohio 43210, and Consulting Engineer, 5540 Abington Drive, Newark, California 94560
Total gaseous mercury in headspace air was measured for enclosed concretes dry curing at 40 °C for intervals of 2, 28, and 56 days. Release of mercury was confirmed for ordinary Portland cement concrete (OPC) and three concretes in which class F fly ash substituted for a fraction of the cement: (a) 33% fly ash (FA33), (b) 55% fly ash (FA55), and (c) 33% fly ash plus 0.5% mercury-loaded powdered activated carbon (HgPAC). Mean rates of mercury release (0.10-0.43 ng/day per kg of concrete) over the standard first 28 days of curing followed the order OPC < FA33 ≈ FA55 < HgPAC. The mercury flux from exposed surfaces of these concretes ranged from 1.9 ( 0.5 to 8.1 ( 2.0 ng/m2/h, values similar to the average flux for multiple natural substrates in Nevada, 4.2 ( 1.4 ng/m2/h, recently published by others. Air sampling extending for 28 days beyond the initial 28-day maturation for OPC, FA55, and HgPAC suggested that the average Hg release rate by OPC is constant over 56 days and that mercury release rates for FA55 and HgPAC may ultimately diminish to levels exhibited by OPC concrete. The release of mercury from all samples was less than 0.1% of total mercury content over the initial curing period, implying that nearly all of the mercury was retained in the concrete.
Introduction An estimated 12 Mt of coal fly ash from electrical power generators in the United States is incorporated into structural concretes and grout (1). Fly ash, which replaces some of the cement normally used in concrete, enhances the material properties of both freshly prepared concrete and hardened concrete and reduces CO2 emissions associated with cement production and use (2). Furthermore, the substitution of inexpensive fly ash for cement reduces the cost of concrete. Fly ashes produced from bituminous and subbituminous coals contain both Hg(0) and Hg(II) at combined total average Hg concentrations of 0.1-0.2 mg/kg (3). The oxidation state of Hg associated with fly ash largely depends on combustion conditions within individual boiler furnaces. Mercury is volatilized and converted to elemental Hg in the high * Corresponding author phone: (614) 292-4061; fax: (614) 2923780; e-mail:
[email protected]. † The Ohio State University. ‡ Consulting Engineer. 10.1021/es050026w CCC: $30.25 Published on Web 06/15/2005
2005 American Chemical Society
temperature combustion chambers of coal boilers. A fraction of this Hg subsequently is oxidized as the flue gas cools, thus converting some gaseous elemental Hg to gaseous oxidized Hg primarily in the forms HgCl2 and HgO (4, 5). The US EPA Clean Air Mercury Rule will reduce mercury emissions from coal-fired power plants. A technology for attaining significant reduction in emitted Hg relies on injection of powdered activated carbon (PAC) into the flue gas stream and would add small quantities of powdered activated carbon containing Hg (HgPAC) to fly ash particulates commonly captured from flue gas. As this technology is implemented, current initiatives for reuse of coal combustion byproducts may be affected. PAC injection will increase levels of both carbon and Hg in fly ash. Previous work explored the possibility of Hg release from unaltered fly ash (6). However, an understanding of the fate of Hg in concrete is needed to enable future applications of fly ash within the framework of evolving regulations. Concrete is a porous material, and Hg bound to either fly ash or PAC ultimately may be released to the atmosphere after concrete is poured. Furthermore, during the mixing, pouring, and initial stage of curing, the temperature of concrete can increase up to 40 °C (7). This temperature elevation, while relatively small, may increase the diffusion and subsequent vaporization of Hg from concrete. The research described herein provides the first report of mercury release into air from concretes prepared with fly ash and mercury-loaded PAC. Specialized techniques created for air sampling above dry curing concretes enabled measurements of mercury that contribute to a new understanding of relationships between mercury mass release rates and concrete composition. Furthermore, the potential influences of curing chemistry and microstructural changes have been noted as important factors in the release of mercury from concrete matrixes that contains fly ash and PAC.
Experimental Methods Concrete. Four types of concrete prepared for this study included ordinary portland cement concrete (OPC) and three concretes in which class F fly ash was substituted for a fraction of the cement: (a) 33% fly ash (FA33), (b) 55% fly ash (FA55), and (c) 33% fly ash plus 0.5% Hg-loaded powder activated carbon (HgPAC). The concrete formulation followed the recommendation by Cannon (8) for fly ash concretes; i.e., 13% binder (fly ash and/or cement), 6% water, 30% sand, and 51% course aggregate. A water-to-binder mass ratio of 0.46 was maintained. An air entrainment admixture (AEA), MicroAir 100 (Degussa, Cleveland, OH), was used to ensure 5-6% air entrainment. Addition of HgPAC in an amount equal to 0.5% of the cement maintained the carbon level just under 6%, as prescribed by the American Concrete Institute (ACI) for applications of fly ash in concrete (2). Major ingredients used in preparation of concrete were high purity (18 MΩ-cm) water (Millipore, Billerica, MA), commercially available type I portland cement, generalpurpose sand, and limestone aggregate. Class F fly ash with a loss on ignition (LOI) of 4.23% ( 0.06% (n ) 3), organic carbon content ranging from 3.8% to 4.4%, and 89.4% of major oxide composition (SiO2 + Al2O3 + Fe2O3) was used in the concretes. The PAC was manufactured by Norit Americas (Darco FGD) specifically for removal of Hg from hot flue gases produced by coal combustion. The specific surface area of fly ash and PAC was determined to be 3.0 m2/g and 481 ( 7 m2/g, respectively, using the BET (Brunauer-Emmett-Teller) technique (FlowSorb 2300II, Micromeritics, GA). The specific surface area of fly ash is VOL. 39, NO. 15, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
5689
FIGURE 1. Air sampling apparatus. within the range 0.45-9.44 m2/g determined by other researchers (9). HgPAC was prepared in the laboratory by exposing a column of 100 g of PAC to a flow of air that passed over metallic Hg for approximately 4 weeks. Experiments. Mercury release during early curing (days 1 and 2) in addition to standard maturation (28 days) and extended curing (days 29 through 56) were investigated by a purge-and-trap approach (Figure 1). For each type of concrete, a freshly mixed batch was divided into two approximately equal portions. Each portion was transferred into a clean HDPE container that finally was sealed by a HDPE gastight lid. Containers of concrete then were moved into an environmental chamber that was maintained at 40 ( 1 °C for curing. During the extended curing (days 29 through 56), the temperature in the chamber unexpectedly drifted from 40 to 23 °C. Airflow of 0.35 L/min through a 2-3-L headspace above the concretes was induced by a microprocessor-controlled AirCheck 2000 air sampling pump (SKC, Eighty Four, PA). A large iodated carbon (IC) trap (0.63 g carbon) (Frontier Geosciences, Seattle, WA) was used to remove Hg from air pulled into each concrete container. Small IC traps (0.36 g carbon) attached at the outlet ports were used to collect Hg species released from concrete (10). Mercury Measurements. The concentration of Hg in each batch of concrete was estimated from Hg concentrations in ingredients. The Hg contents of cement, sand, limestone aggregate, MicroAir 100, Hg-loaded PAC, and high-purity water were determined by cold vapor atomic fluorescence spectrometry (CVAFS). Cold vapor atomic absorption spectrometry (CVAAS) was used to determine the Hg concentration of fly ash (11). Prior to all measurements, samples of solid materials were digested in high-purity nitric acid by microwave heating within sealed PTFE vessels. Determinations of Hg in the National Institute of Standards and Technology (NIST) Standard Reference Material 1641d fell within 93.5-98.7% of the certified concentration. The Hg collected on the IC traps was determined by dual amalgamation preconcentration and CVAFS at Frontier Geosciences (11) after digestion in a mixture of 70% nitric acid and 30% sulfuric acids (v/v). The detection limit for Hg on IC traps ranged from 0.1 to 0.7 ng/trap. Recoveries for 50 or 100 ng spikes on IC traps ranged from 95.2% to 102%. Control Experiments. Mercury release experiments for each type of concrete were conducted, minimally, in 5690
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 15, 2005
triplicate. In each experiment, a sample blank, that is, an empty container sampled alongside containers holding concrete, enabled estimates of the “system contribution” to the total Hg collected. Each sample blank was subtracted from the total Hg collected for a corresponding concrete sample to provide a corrected Hg value for that sample. Quantities of Hg collected for system blanks ranged from 0.96 to 7.29 ng for the 2-day curing interval and from 13.4 to 41.0 ng for the standard 28-day maturation. Selected large and small IC traps were analyzed for partitioning of Hg between the upstream and downstream segments (within each trap structure). The mass ratio for Hg at a downstream segment to a corresponding upstream segment was less than 0.04 and 0.08% for large and small IC traps, respectively. Therefore, negligible breakthrough occurred for the IC traps used in this study. No measurable contribution of Hg from the air pulled into the containers took place, and all collected Hg emissions from concretes remained on the small IC traps.
Results Mercury in Concrete. The Hg concentration in each component used to make concretes is shown in Table 1. Importantly, these ingredients had Hg concentrations similar to those reported in other published studies. The Hg concentration in fly ash was 0.119 mg/kg, which was well within the range of concentrations in fly ash generated from combustion of eastern coals, that is, 0.02-4.2 mg/kg (12). The Hg concentration in the loaded PAC was 19.9 mg/kg, a level of Hg loading that falls within the range observed in full-scale testing of PAC at coal-fired power plants (13). Although Hg concentrations were relatively low in sand and coarse aggregate, the large masses of these components in the prepared concrete, relative to cement, fly ash, and other ingredients, made them significant contributors to total Hg in concrete. The major sources of Hg in each batch of concrete were Hg-loaded PAC, fly ash, sand, and coarse aggregate (Table 2). Cement, water, and air-entrainment admixture contributed little Hg. Mercury Release. The variability of measured Hg on IC traps was high (14-93% RSD) for samples collected over the first 2 days of curing. The small quantities of Hg trapped (3.54-17.8 ng, blank corrected) were not conducive to good precision, because of their proximity to the method detection
TABLE 1. Mercury Concentrations in Concrete Components concentration mg/kg sand this study literature a
0.0072 0.02-0.1
cement 0.0028 0.06 (19)
mg/L
coarse aggregate 0.0031 0.005-0.46 (20)
fly ash
HgPAC
0.119 0.1-1.0
water 10-7
