A wetland was constructed near Norton, VA by Westmoreland Coal Company to treat acid mine drainage (AMD) from an inactive coal refuse pile. The AMD had an average inflow pH of 7.0, and average inflow concentrations measuring 4 mg/L total Fe, 3 mg/L total Mn and 450 mg/L dissolved sulfate. An 18 month field study of water quality improvement and sulfate-reducing bacterial (SRB) populations revealed that the wetland was effectively treating the AMD. Iron and Mn both met compliance standards set by the EPA requiring an instream Fe concentration of 3 mg/L and an instream Mn concentration of 2 mg/L. SRB averaged 8.7 x 10⁴ through the 18 month study period as determined by the Most Probable Number (MPN) method. The concentration of sulfate was decreased by an average of 360 mg/L as the AMD passed through the wetland.

In a separate laboratory study, the effect of carbon addition, pH and Fe concentration on microbial sulfate reduction and the subsequent precipitation of Fe and Mn was determined in mesocosms built to simulate a wetland. Mesocosms were constructed with plexiglass sheets and measured 6" x 6" x 24". Each mesocosm was filled with a 4-inch layer of limestone gravel beneath 17 inches of weathered pine bark mulch. A perforated PVC pipe was installed within the limestone layer to act as an underground drain. With mulch as the only source of available carbon, a 15% decrease in total sulfate concentration occurred in AMD containing initial concentrations of 1500, 750 and 375 mg/L sulfate. The population of SRB averaged 10⁵/ g dry mulch. The addition of 300 mg/L carbon as lactate resulted in an a 3 log₁₀ increase in SRB population. Following the addition of carbon as lactate, the concentration of sulfate decreased 95%. Total Fe decreased 90% from inflow concentrations prior to the addition of lactate, and decreased 96% following the addition of lactate to the AMD.

The effect of varying the influent pH of AMD was studied using wetland mesocosms, and a pH of 3.5 adversely affected microbial sulfate reduction and water quality improvement. Populations of SRB decreased by 3 log₁₀ from an initial population of 10⁸ SRB/g dry mulch. Iron and Mn concentrations decreased 70 and 37% respectively. Hydrogen ion concentration increased to 7.0 and above when inflow pH was 4.5 and 6.0, but increased to an average of 6.4 when inflow pH was 3.5.

The effect of different concentrations of Fe within AMD was investigated using wetland mesocosms, and total inflow Fe concentrations of 155 and 301 mg/L resulted in a greater percent decrease in sulfate concentrations than at a lessor Fe concentration averaging 85 mg/L. Total Mn decreased 12% at an inflow Fe concentration of 85 mg/L, and decreased 43% at an inflow Fe concentration of 301 mg/L. The results generated from both the analysis of the Pine Branch wetland and the laboratory mesocosm experiments demonstrate that subsurface flow constructed wetlands are a viable form for treatment of AMD. 2197820b-4775-4425-b667-55393f34b513,"This thesis deals with the deliberate insertion of nonlinear elements in second-order linear control systems for the purpose of improving their transient response. The main body consists of a method of obtaining a desired step response by placing a nonlinear computer in the forward loop. This computer fixes the system trajectory in the phase plane by determining the required output velocity for the error present at any time. An inner control loop adjusts the output velocity to agree with the computed signal in an extremely short time, thus giving a very close agreement between actual and desired responses.

Several examples are presented to show the application of this method, and experimental verification is obtained with an analog computer. Areas of future study and practical limitations are discussed in the final sections of the thesis.