COMPLETE CHARACTERIZATION OF PARAMETERS USED IN
RISK ASSESSMENT MODELS FOR HEAVY METAL TRANSPORT
ASSOCIATED WITH FERTILIZER APPLICATIONS IN OREGON
Progress Report
December 2004
Submitted to Oregon Department of Agriculture, Pesticides Division
by:
William Fish and Gwynn R. Johnson
Portland State University
Contacts:
W. Fish [email protected] (503) 725 – 4278
G.R. Johnson [email protected]
(503) 725 - 8710
Executive Summary
This report summarized research activities performed at Portland State University under contract with the
Oregon Department of Agriculture (ODA) related to understanding the behavior of fertilizer derived
metals in Oregon agricultural soils. The goal of the PSU project is to create a model of metal solubility,
transport, and accumulation in agricultural soils that requires a minimal number of measured physical and
chemical parameters, yet represents a diversity of Oregon soil types and agricultural practices.
In the prior reporting period (January to June 2005) we reported that we had initiated or completed the
following tasks in the proposal timeline:
Task 1. Collected bulk and intact-core soil samples from four trial site locations in Oregon
Task 2. Obtain fertilizer samples and other necessary reagents and supplies (complete).
Task 3. Processed soil samples in laboratory for particles size distribution, pH, and metal content
Task 4. Initiated laboratory-based soil characterizations
In this reporting period (July 2004 through December 2004) we accomplished the following tasks, as
identified for this period in our proposal:
Task 1. Collected additional bulk and intact-core soil samples from the trial site locations
Task 3. Completed laboratory analyses for soil particle-size distribution, pH, and metal content
Task 4. Continued laboratory-based soil characterizations, in particular the acid-base characteristics of
the soil.
The results of this work have allowed us to continue to fully document the parameters we need for our
ultimate goal of a robust model of fertilizer-derived metal behavior. Among our findings were that
•
Soil particle-size distributions were completed, yielding important physical grain-size
characteristics necessary for the modeling of water and metal transport in soils
•
Depth profiles of Cd at the four field sites were obtained revealing that most of the soil Cd is
present in the upper 30-40 cm, probably corresponding to the tillage zone.
•
The concentration and the depth of elevated Cd in the soils did not correlate with the treatment
levels from he first-year experiment; no distinction was possible among the control or any of the
three treatments.
•
Alkalimetric (acid-base) titrations of soils from all sites were completed. The parameters
obtained allow us to accurately model the effects of pH variation on metal behavior in soils.
Background
Fertilizers, agricultural minerals, agricultural amendments, and lime products may contain toxic metal
contaminants that can adversely affect human health and the well-being of livestock and natural
ecosystems. The levels of heavy metals in fertilizers and related materials are thus subject to regulation by
the Oregon Department of Agriculture. Regulators need to balance the benefits of economical fertilizers
with the risks posed by excessive levels of metals in these essential products. Balancing benefits and risks
can be achieved with risk-based standards. Human health risk assessments are a key part of creating
reasonable and prudent regulations for permissible levels of metals in fertilizers.
Although studies in recent years have greatly expanded the base of knowledge for such systems, much of
the work has focused on metals derived from hazardous wastes or sewage sludge application. Also, there
is very little available information pertinent to metal behavior in Oregon soils under a variety of
agricultural practices and climate conditions. Thus it is difficult or impossible to transfer the results of
those studies to the specific problem of metals leaching from fertilizers as they are used in Oregon
agriculture. There is an urgent need for a detailed yet practical study of key heavy metals in the fertilizersoil systems that are relevant to Oregon applications.
OVERALL GOAL AND SPECIFIC OBJECTIVES
Our overall goal is to create a model of metal solubility, transport, and accumulation in agricultural soils
that requires a minimal number of measured physical and chemical parameters, yet represents a diversity
of Oregon soil types and agricultural practices. The model and its supporting data will be used to: 1)
assess the leachability and availability of soluble metals; 2) characterize the potential for long-term
buildup of metals in soils; and 3) identify the rate at which accumulated metals either leach from the soil
or are sequestered via an “aging” process. The specific objectives of this research project are to:
1. Collect intact “undisturbed” core samples along with corresponding conventional grab
samples of soils from trial sites located in Oregon, coordinating with K. Anderson (OSU) and
representatives from ODA.
2. Characterize the physical properties of undisturbed soils using advanced column techniques
that reveal the role of natural heterogeneity in soil structure and chemistry.
3. Analyze soil grab samples for conventional physical/chemical characteristics such as
porosity, mineralogy, cation exchange capacity, organic carbon content and extractable metal
oxides.
4. Identify the metal sorption/desorption properties of the various soil samples over a wide
range of metal concentrations, pH, and for relevant (target) toxic metals.
5. Characterize the importance of rate-limited mass-transfer processes of sorption and
desorption with a special emphasis on the aging associated with long-term heavy metal
loaded soils that may potentially effect the eventual leaching of metals.
6. Create a practical model of metal-soil interactions based on parameters obtainable from
conventional soil characterization methods.
7. Verify the diagnostic utility of the model with column studies of undisturbed cores that
bridge the gap between conventional lab studies and actual behavior of metals in the field.
Research Accomplishments in this Reporting Period
1. Completed grain –size distribution determinations for all soil samples. We now have the necessary
size data such ad the d10 and d90 values that are essential for quantifying soil particle size heterogeneity,
hydraulic conductivity, and predicted solute dispersion coefficients. Typical grain size distribution data
are shown below in Fig. 1
100
0.300,
87.164
Percent Finer
80
60
0.215,
59.858
0.150,
44.997
40
0.075,
20
21.720
0.053, 7.303
0.045, 4.918
0
1
0.1
0.01
Grain Size (mm)
Fig. 1: Grain Size Distribution for Columbia Basin soils.
2. Completed collection of additional bulk and intact-core soil samples from the trial site locations,
including cores for characterizing depth distribution of Cd. Sites that we were unable to sample last
fall due to dry soil conditions were sampled and sets of bulk, intact, and depth cores collected. The Intact
cores are archived for future experimentation. During this reporting period we analyzed the depth cores
(1" diameter cores taken to a depth of 50 cm) for the distribution of soil-bound Cd with depth. Depth
profiles of Cd from the four field sites revealed that most of the soil Cd is present in the upper 30-40 cm,
probably corresponding to the tillage zone (Figs. 2 and 3).
Also, no significant variation in Cd levels could be detected among the control and three treatment levels,
probably because of the low levels of Cd in the fertilizer used in the first year. Variations were on the
order of 50% but they did not correlate with treatment levels and the control plots were not statistically
significantly different from any of the treatment levels.
However, useful information can still be gleaned from the Cd profiles. The persistence and
concentrations of Cd in the upper, tillage zone of the field plots, when coupled with the absence of a Cd
signal from the treatments, suggests that the Cd profiles are the result of many prior years of ordinary
fertilization in the course of station experimentation. This implies that there may be a slow but
measurable accumulation of relatively low levels of Cd in these soils. The accumulation in the tillage
zone (down to 20-40 cm) is consistent with a fertilizer-derived source but could also possibly be
explained by atmospheric deposition or other causes.
0.024
Cadmium Depth Profiles for Columbia Basin
and Klamath Falls
0.019
[Cadmium] (mg/kg)
CHB3104
CHA3104
0.014
KH3104
0.009
0.004
-0.001
0-5
5-10
10-15
15-20
20-25
25-30
30-35
35-40
40-45
45-50
Depth (cm)
Fig. 2. Depth variation of Cd in the upper 50 cm of core samples from Klamath and Columbia Basin
0.025
KH1104
KH1204
KHI304
KH2204
KH2304
KH2404
KH3104
KH3204
KH3304
KH3404
KH4104
KH4204
KH4304
KH4404
KH1404
Cadmium Concentration
with Depth at Klamath
[Cadmium ] (mg/kg)
0.020
0.015
0.010
0.005
0.000
A
0-5
B
5-10
C
10-15
D
15-20
E
20-25
F
25-30
G
30-35
H
35-40
I
40-45
J
45-50
K
50-55
L
55-60
M
60-65
Depth (cm)
Fig. 3. Variation in Cd with depth in all samples plots at Klamath station, showing no significant
difference among treatment levels.
3. Alkalimetric (acid-base) titrations of soils from all sites were completed. The parameters obtained
allow us to accurately model the effects of pH variation on metal behavior in soils. Toxic metals bind to
surfaces by several mechanisms. Metals exhibit typical cation-exchange behavior on clay minerals; ions
such as Cd2+ and Pb2+ are attracted to layer silicates by electrostatic forces. However, these metals must
compete with more abundant major cations for the available exchange sites, so a large degree of metal
partitioning onto exchange sites is not expected. Sorption of metals on metal oxides (such as hydrous
ferric oxide) is generally far more important and can be expressed as a surface-complexation reaction of
the type:
Fe--OH] + Cd2+ = Fe—O--Cd]+ + H+
Note the involvement of the hydrogen ion in the reaction, meaning that the reaction is pH dependent. In
this stage of the research we focused on obtaining detailed information about the acid-base (proton)
reactions of the soil samples.
Alkalimetric Soil Titrations
WBT-1
A- = 10.58pH - 50.6
65
60
55
A-, ueq/g soil
50
45
40
35
30
25
20
15
10
5
0
2
WBT-1
R = 0.998
WBT-2
WBT-2
A- = 11.83pH - 57.5
HBT-1
2
R = 0.992
HBT-2
Linear (WBT1)
Linear (HBT1)
Linear (WBT2)
Linear (HBT2)
HBT-1
A- = 3.37pH - 17.5
2
R = 0.991
HBT-2
A- = 3.80pH - 18.4
2
R = 0.991
-5
4.00
6.00
8.00
10.00
pH
Fig 4. Alkalimetric (acid-base) titrations of Willamette and Hermiston soil samples. Note the
consistency within replicates of a station site but the significant different between the stations.
Apparent pKa Distribution with pH
10
9
HBT-1
pK-eff
8
WBT-1
7
HBT-2
WBT-2
6
5
4
4.5
5.5
6.5
7.5
8.5
pH
Fig.5 Apparent pKa distribution with pH for site soils.
Electrostatic Interaction Plots
10
9
HBT-1
WBT-1
HBT-1
y = 2.34x + 6.37
HBT-2
R = 0.993
2
WBT-2
log K'
8
7
6
HBT-2
y = 2.66x + 6.01
Linear (HBT1)
Linear
(WBT-1)
Linear
(WBT-2)
Linear (HBT2)
R2 = 0.989
WBT-1
y = 3.01x + 5.90
R2 = 0.989
WBT-2
y = 2.35x + 6.06
5
2
4
0.00
R = 0.940
0.20
0.40
1 - alpha [A-]
Fig.6 Electrostatic interaction plots for site soils.
0.60
0.80
Henderson-Hasselbach Plots
10
9
8
pH
HBT-1
7
WBT-1
6
5
4
-1.00
0.00
1.00
2.00
3.00
log alpha/(1 - alpha)
Fig. 7 Henderson-Hassebach plots of the alkalimetric titrations for site soils.
A-, ueq/g soil
Alkalimetric Soil Titrations
with Cd Added
65
60
55
50
45
40
35
30
25
20
15
10
5
0
-5
-10
4.00
KBT-1
y = 10.052x - 50.429
2
R = 0.9456
KBT-1+Cd
y = 6.2654x - 30.584
R2 = 0.9946
6.00
8.00
10.00
KBT-1
KBT-1+Cd
Linear (KBT-1)
Linear (KBT-1+Cd)
pH
Fig 8. Effect of added Cd on the pH behavior of the soil, providing alternative information
about proton-metal interactions.
Next Steps in the Research Plan.




Measure Cd adsorption explicitly as a function of pH
Work up model of surface-site/proton interaction (surface acid-base)
Combine with adsorption isotherms previous described to create comprehensive model
on metal binding for all soils
Look at variation at and among sites
PRINCIPAL INVESTIGATORS
Biographical Sketches
Dr. William Fish
Dr. Fish is an Associate Professor at Portland State University with joint appointments in both
the Department of Civil & Environmental Engineering and the Environmental Sciences &
Resources program. Dr. Fish’s expertise focuses on the interactions of chemical and physical
processes in soils and aquatic environments, and the ways those interactions influence the
behavior of pollutants. Dr. Fish uses both laboratory experimentation and computer simulations
to better characterize how contaminants such as heavy metals move through the environment. He
has published a number of articles on metal-humate interactions, surface-complexation modeling
of metal adsorption of oxide surfaces and analytical methods for metals in soils and sediments.
One recent project focuses on ways in which protective capping materials for contaminated
sediments can be chemically modified to enhance their ability to trap and retain toxic metals. In
related work, Dr. Fish has collaborated on a study of the redistribution of metals during episodic
disruption of sediments in dynamic estuarine systems. Another project centered on modeling of
fate and transport of PCBs and toxic metals in the Columbia Slough. The goal in all of these
projects is to use scientific understanding to foster better decision-making in the realm of
environmental management.
Dr. Gwynn R. Johnson
Dr. Johnson is an Assistant Professor at Portland State University with joint appointments in
both the Department of Civil & Environmental Engineering and the Geology Department. Dr.
Johnson’s expertise focuses on contaminant transport and fate in porous materials. Specifically,
Dr. Johnson has investigated and published on the effects of physical and chemical
heterogeneities on contaminant transport through undisturbed vs. homogenized aquifer cores.
Her interests encompass many fields of study, including environmental chemistry, soil and water
sciences, groundwater hydrology, water resources, and environmental engineering. Much of her
research has focused on the multiple physical and chemical mass transfer processes influencing
the transport, fate, and remediation of contaminants in the subsurface. Additionally, her research
has involved characterization/modeling of flow and transport through porous media, including
homogeneous and naturally heterogeneous materials.