Search Documents

By Timothy P. Smirnoff, Tracy K. Lundin

INTRODUCTION The City of Phoenix, Arizona is located in the Phoenix Basin watershed which is prone to flash flooding several times each year. The runoff from these storms moves from the mountains in the north to the Salt River in the south. An extension of Interstate Highway 10 (1-10) will pass through the city east to west to form an Inner Loop through the downtown business district. PROJECT DESCRIPTION The I-10 Inner Loop extension will be a depressed roadway which would intercept the surface drainage of the Phoenix Basin and separate the watershed into two regions. To prevent flooding of the depressed roadway, runoff storm water collected from the northern half of the drainage area will be conveyed beneath the roadway to its natural outlet at the Salt River. A conveyance system composed of pressure tunnels beneath the city, shown in Figure 1, was proposed to achieve this goal . Table A presents the prominent features of each of the tunnels. This paper will describe geologic and engineering parameters specific to Phoenix which were used in the design of a series of tunnels to convey storm water from the I-10 Inner Loop to the Salt River. The design of tunnel initial and final support is described along with support and construction methods used by the Contractor.

Jan 1, 1985

By Paul Zick

The Port Authority of Allegheny County North Shore Connector (NSC) Project involves TBM excavation of twin 690 m (2,240 lf) precast segment lined light rail tunnels through soil and rock under the Allegheny River and a narrow, downtown city street. The $435 M project is funded 80% ($348 M) by the Federal Government, 17% ($72.5 M) by the State of Pennsylvania and 3% ($14.5 M) by Allegheny County. Along the tunnel alignment are a historic building on shallow foundations, modern buildings on deep foundations, and one building constructed directly over the alignment. The project includes construction of twin 369 m (1200 lf) cut and cover concrete box tunnels and an underground station shell. This paper provides an overview of general site conditions, current project status and construction challenges for the Twin Tunnels and North Shore Light Rail Station.

By D. P. Nicholson

The current phase of office development at Canary Wharf in the London Docklands includes twelve multi-storey buildings varying from eighteen to forty stories in height. The area of Canary Wharf South comprises five multi-storey buildings with low level in fill podium and garden areas. Three of these multi-storey buildings have "pile-cap" designs where the entire building load is applied to the piles. The remaining two buildings have "piled-raft" designs where allowance is made in the design for part of the building load to be transferred to the raft formation. Using a piled-raft design approach has enabled savings to be made on the number of piles used below the raft. This paper describes the calibration analysis undertaken to validate the design method. The ground conditions at the Canary Wharf site are typically fill and recent gravel deposits over the overconsolidated clay and sand of the Lambeth Group. Below these strata is the Thanet Sand. The Lambeth Group provided the founding layer for the raft and contributed to the pile shaft resistance. The Thanet Sand stratum provided the bearing layer for the 1.5m diameter base grouted bored piles. The piled-raft design required an assessment of the building load transfer through the raft to the formation layer. This load transfer led to significant increases in effective stress in the soil below the raft and thence a corresponding increased the capacity of the piles. This interaction of the raft with the prepared formation contributed to between 27% and 45% increase in total vertical load carrying capacity of the piled-raft substructure compared with the pile-cap design method. The development of the piled-raft design was carried out in three main phases: ? Development of the approach and validation analyses. ? Design work on various schemes resulting in detailed design packages. ? Validation analyses using full 3D finite element analysis (presented herein). The validation of the piled-raft design approach was initially carried out against published case histories for piled-raft interaction. To improve on the validation work a full 3D finite element analysis of a simplified pile raft and superstructure (which contributes stiffness to the raft) was undertaken; the analysis included only a quarter of the structure due to assumed symmetry of the structure and vertical loading. This 3D validation analysis is presented herein. The analysis was carried out using the Oasys LS-DYNA finite element program. Prior to carrying out the analysis of the piled-raft an initial finite element analysis was carried out to back analyse a fully instrumented field pile test providing accurate modelling of a single pile performance. This back analysis allowed confidence in the soil parameters used in the 3D model; comparisons of shaft and base resistances between the pile test and the FE model calculations for the single pile are presented. The parameters provided in this work were then carried into the full 3D model of the piled-raft. The 3D model of the piled-raft structure included 18 piles (12 full piles and 6 half piles on an axis of symmetry of the model).

Jan 1, 2002

By I. A. Costa, Y. Fouquet, F. Barriga

"Lucky Strike and Menez Gwen are two hydrothermally active sites located along the border of the Azorean plateau, southwest of the Azores islands where the alteration processes affecting basaltic rocks are prominent (Costa et al., 2003). Here we report on a slabby brecciated rock, abundant at both Lucky Strike and Menez Gwen, composed mainly of basaltic fragments cemented by hydrothermal precipitates from warm fluid circulation system. Research reveals that hydrothermal alteration takes place mainly by replacement processes, in which silicification is dominant (Costa et al., 1997 and Costa, 2014).Petrographic observations, including scanning electron microscopy (SEM), microprobe analysis, and mass balance studies made on basaltic fragments within the breccia show that silicification is one of the main processes resulting from hydrothermal fluid circulation. This alteration process leads to complete replacement by silica, in extreme cases, forming an impermeable cap that may contribute to the formation of sub-seafloor sulphide deposits, including examples in the geological past."

Sep 1, 2014

By J. Noll

Many underground metal/nonmetal mines are using enclosed cabs on mining equipment to reduce the exposure of miners to diesel particulate matter (DPM). However, some enclosed cab systems may not be meeting their expected efficiency in capturing DPM, which could be a result of inappropriate filter use, cab leaks, or insufficient pressurization. NIOSH investigated several types of filters for reducing DPM to determine which type had the best capture efficiency. The laboratory results indicated that a MERV 8 (Minimum Efficiency Reporting Value, as defined by the American Society of Heating, Refrigerating and Air- Conditioning Engineers) filter captured about 50 percent of the DPM, a MERV 16 filter collected 96-98 percent of the DPM, and two HEPA (high efficiency particulate air) grade filters captured over 99 percent of the DPM. NIOSH also performed a field study to measure the efficiency of two cabs at a limestone mine. The cab filtration systems of a loader and haul truck were found to be over 90 percent effective in removing DPM except under certain operating conditions when the efficiency was as low as 41 percent. Evidence seems to suggest that these conditions were when a window was open in the cab.

Jan 1, 2011

?Classification System Efficiency (CSE)? is a simple, powerful concept which can be used to measure and improve ball mill circuit efficiency. This paper describes six plant case studies which demonstrate the use of CSE to diagnose and/or improve grinding circuit performance, with related improvements in circuit efficiency ranging from 3 to 30 percent. The increased effective mill grinding energy is recaptured from energy that was going into over grinding, which can greatly benefit downstream mineral separation. The achievable CSE for given circuit design/operating variables is described, providing every plant metallurgist with the means to determine the potential for increasing their particular grinding circuit?s efficiency through increased CSE.

Jan 1, 2012

By A. C. Halferdahl

The Granby ores may be divided into three groups or classes. The No. I ore is a heavy pyrite, chalcopyrite, pyrrhotite, sulphide carrying some free silica, and calcite. No. 2 ore carries the same general group of sulphides as does No. 1, but in a lesser amount. The gangue is an altered greenstone. No. 3 ore has a smaller amount of sulphides than either of the others. Analyses of these ores is as follows: Cu. Si02 Fe. CaO S. Al 203 MgO No. I. Per Cent. 1.8 17.5 33.0 5.7 34.5 2.3 1.3 No. 2 Per cent. 2. 0 35. 6 23. 9 4. 3 I .5. 0 9. 2 6 .. 0 No. 3, Per cent. 1 . 5 40. 9 21 . 7 3. 0 I 1 . 3 11 . 0 5 . 4 The sulphide minerals in No. 2 and No. 3 ores occur largely in bands in the gangue, and for these portions comparatively coarse grinding serves to release the gangue. Drill cores show a certain dissemination of sulphides in portions of the rock, so that grinding to 5 per cent. + 65 mesh with 65 per cent.-200 mesh liberates the sulphide minerals very completely. The association of the sulphides is fundamental. The pyrite was probably deposited first and was followed by a portion of the chalcopyrite, or these two portions of the sulphide minerals were deposited together.

Jan 1, 1924

By H Kurth

Geoscan is used for multi-elemental analysis of conveyed bulk materials in real time. It has been used for coal analysis since the 1990s, in the cement industry, and since 2003 for measuring mineral ore quality. Geoscan-M has proven performance and demonstrated paybacks of weeks or months, in iron ore, manganese, copper, zinc-lead and phosphate industries. Geoscan-M measures continuously through the full bed depth and is unaffected by belt speed, particle size, layering and dust levels. Increments of two minutes (typically one to five minutes) are accurately measured and provide the plant operators with tonnage weighted results of each element within seconds.A Geoscan installed after crushing is able to measure quality shortly after mining. The analysis results can be used to divert the increments based on composition, such as diverting waste for disposal and diverting marginal material to a stockpile for processing at a later date. The bulk sorting process allows different qualities to be placed into relevant stockpiles for subsequent blending. Material meeting product DSO (Direct Shipping Ore) quality can bypass a beneficiation circuit to prevent unnecessary treatment costs. An analyser in a blending application enables proportioning of different quality feed materials from multiple stockpiles to achieve a desired compositional consistency for process feed. The parameter used for decisions may be; a) primary metal content, e.g. Cu, Fe, P, etc. b) deleterious component content, e.g. Mg as an indicator of talc content, c) process parameter content, or a combination of these. Process parameter content may be the elemental ratio between critical components, e.g. Cu:S ratio, Fe:Si ratio, basicity, Lime Saturation Factor, etc. or the content of an “indicator” element, e.g. Si as an indicator of material hardness or grindability.Measurement of the composition in real time enables prompt feedback to the mining operators and also feed forward to the process operators. Compositional variability can be critical in the process and advance warning enables appropriate reactions prior to processing. This may include modifying the feed rate, changing the feed blend (through additives), or modifying reagent dosing in anticipation of a change in feed quality. Benefits include, but are not limited to; reducing waste processing, improved feed grade stability, reducing energy, water and grinding media consumption, producing less tailings, increasing process efficiency, higher metal recovery, better plant utilisation, reducing plant downtime, better utilising mined material, all achieved using the same process plant throughput capacity.CITATION:Kurth, H, 2015. Optimising plant feed quality and process performance using Geoscan elemental analysis, in Proceedings MetPlant 2015, pp 373–380 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Sep 7, 2015

By Bruce Notthup

Using a Three Dimensional Laser Profiler, persomrel from Austin Powder were able to determine drilling patterns for use in underground mine Room and Pillar blasting operations, The faces had got out of alignment, causing concave faces across the 100 feet wide, 28 feet high tunnels. This was leading to poor drilling and poor blasting results. By profiling the faces, modifications could be made to the drilling pattern to help gradually straighten the faces. The Laser Profiler was also used to check the accuracy of collaring the holes and also perform post blast surveying of the positions of the bootlegs. This allowed recommendations to be made to improve drilling operations at the face. In addition, other volumetric calculations could be performed using the systems’ surveying software, all a lot more easily and quickly than by standard methods.

Jan 1, 1994

By AIME AIME

General Local Committee.-John Fritz, Chairman; A. B. de Saulles, Vice-Chairman.; R. M. Bird, Treasurer; Henry S. Drinker, Secretary; Robert H. Sayre, E. P. Wilbur, Charles M. Dodson, Abraham S. Schropp, E. M. McIlvain Archibald Johnston, A. B. Fichter, C. G. Convers, W. A. Wilbur, Garrett B. Linderman, Frederick Conlin, George Pettiuos, Albert Brodhead, Garrett L. Hoppes, M.B. Cutter, H. J. Seaman, C. A. Matcham, B. F. Fackenthal, Jr., A. L. J. Queneau, Geo. R. Elder, Eldridge Wilbur, William H. Chandler, Mansfield Merriman, Joseph F. Klein, William S. Franklin, William Esty, Joseph AN'. Richards, Howard Eckfeldt, John D. Irving. Transportation Committee.-H. J. Seaman, C. A. Matcham, B. F. Fackenthal, Jr., A. B. Fichter, A. L. J. Queneau, Henry S. Drinker. Hotel Committee. -Garrett B. Linderman, Charles M. Dodson, W. A. Wilbur, Garrett L. Hoppes, Albert Brodhead. Steering Committee.-William S. Franklin, William Esty, John D. Irving, R. M. Bird, the members of the Tau Beta Pi Society of Lehigh University. Committee on Meetings - William S. Drinker, Joseph W. Richards, Howard Eckfeldt. Entertainmment Committee.-W. A. Wilbur, E. M. McIlvain, Archibald Johnston, A. B. de Saulles, G. G. Convers,.A. L. J. Queneau, Garrett B. Linderman, Frederick Conlin, George Pettinos, Garrett L. Hoppes, Henry S. Drinker, William S. Franklin, Eldridge Wilbur, Robert M. Bird, Secretary. The opening session was held at 8.30 p.m., Wednesday, February 21, 1906, in Gymnasium Hall of Lehigh University, which had been tastefully decorated with flags and bunting. Dr. Henry S. Drinker, President of Lehigh University, and Secretary of the General Local Committee, called the meeting to order, and welcomed the Institute to the Bethlehems, where its first meeting, after its organization in May, 1871, at Wilkes-Barre, was held in August, 1871.*

May 1, 1906

By J. A. Finch, A. R. Laplante, R. Espinosa-Gomez

"The Niobec concentrator rejects up to /5% of the Nb as slimes (-10u.m). The recovery of Nb from these slimes by column flotation is being studied. An essential condition for selective flotation is slimes dispersion. A particle size analyzer, the Sedigraph 5000D, was used to monitor dispersion in plant waters with combinations of organic and inorganic dispersants and flotation reagents. Flotation reagents were particularly detrimental to dispersion with sodium silicate. The best dispersion conditions were selected for column flotation tests. These tests revealed that reagent combination had a pronounced effect on froth stability. Careful frother selection achieved good froth stability under conditions of good dispersion. Plant water increased the required dispersant concentration causing mineral depression.IntroductionSlimes (-10 u.m) rejected at Niobec contain up to 15070 of the over-all Nb content . Recovery of Nb from these slimes would significantly increase over-all recovery. A suggested recovery route involves two phases: (1) carbonate flotation (to remove most of the carbonates) depressing silicates and pyrochlore (as currently done at Niobec); and (2) selective (direct) pyrochlore flotation from the silicates. Phase 2 has been addressed and several routes to achieve selective flotation are available. Phase I was attempted with the selective reagent scheme used at Niobec but a considerable amount of pyrochlore floated with the carbonates due to poor slimes dispersion."

Jan 1, 1987

The 18th annual American Mining Hall of Fame Awards Banquet sponsored by the Mining Foundation of the Southwest will be held on Saturday, Dec. 2, 2000 at The Westin La Paloma in Tucson, AZ. The 2000 inductee, guest of honor and featured speaker at the banquet will be Ronald C. Cambre, chairman and chief executive officer of Newmont Mining. He will be honored for his role in positioning Newmont as an innovative and key player in the gold mining industry.

Jan 1, 2000

By Brian Flintoff

"When I was first asked to prepare this paper and presentation in late June, I thought it might be fun to ""step back and paint the bigger picture."" Having attempted to do this, fun is not the first descriptor that now comes to mind, and speaking dyslexically, I've had 'nuf.'Unlike Nostradamus, or even Orwell, I've discovered that I'm ""clairvoyantly challenged"" and not even the psychic hot-line was of much help, at least with respect to process control! I am somewhat comfortable speculating on evolutionary change, but the radical developments that will arise from revolutionary changes are well beyond my mortal capabilities. Even with evolutionary change, I'm reminded of the lessons of time series analysis which show the prediction error to grow exponentially as the horizon moves into the future. As Macluhan is reported to have said: ""Predicting the future based on the past is a little like driving your car while looking in the rearview mirror!""Having honestly presented my qualifications, allow me to try to look at process control and to offer some conjecture on where we're probably headed. In the end, it is said that: ""The plant of the future will have only two employees, a man and a dog. The dog will be there to keep the man from touching the equipment.1""IntroductionA little less than a decade ago Andy Mular and I co-edited a monograph entitled ""A Practical Guide to Process Control in the Minerals Industry."" The work was commissioned by the senior management of a consortium of Canadian mining companies, since it was widely perceived that control was an area of great economic potential, but one that had not been effectively tapped. The concern was that if the industry could not soon deliver on the ""potential"", the term may well have appeared in the epitaph of what was, to these men, a very attractive investment opportunity. I think it is fair to say that over the past few years process control has clearly established itself as an essential technology in plant operation and optimization. That isn't to say that we get it all-right, all-the-time, but, the circumstances are much different than they were. In the eyes of many, process control has attained the distinction of being considered a mature technology, even though there are still countless avenues for exploration."

Jan 1, 1998

Thermal method of diagnostic as a nondestructive monitoring is based on data logging of infrared radiation from heated surface or from the associated thermal field. Monitoring of gearcases and motors hidden faults, such as, bearing degradation (deep-seated wears and tears, fretting wear, spalling and jamming solids of revolution) is closely related to high variations in temperature. The use of portable thermal imaging devices provides considerable advantages due to: easy remote access to the unit under test even when it is operating; minimum time required to obtain information and an opportunity for a real-time diagnosis of the tested subject's condition.

Dec 6, 2010

By Jueun Lee, Jong-Hwa Yoon, K. Rider, Bernard J. Martin

Whole body vibration is known to affect movement accuracy [1], however little is known about changes in the organization of movement and movement strategies used to limit the influences of perturbations. The specific aim of this work is to analyze the motion and coordination of upper body segments of seated operators performing reaching tasks under whole-body sinusoidal vibration exposure and simulated vehicle ride motion. The long-term objective is to model reach coordination and predict the dynamic behavior of the upper body motion under vehicle vibration exposure. Method The reach task consisted of pointing with the right hand index finger to targets located on touch screens placed in front of the subject, 45° overhead and 90° to the right in the mockup cabin of an HMMWV placed on a 6 DOF ride motion simulator. The task was performed under stable (no vibration) and vibration (sinusoidal vibration or simulated ride motion) conditions. A motion capture system was used to record kinematic data of reflective markers to recreate body link trajectories. Joint angles (torso, shoulder and elbow; Figure 1) were then computed using quaternions. Coordination between body links was defined as a) the joints angle-versus-angle relationships between the upper arm and lower arm, and b) the joint motion onset relationships between torso, upper arm and lower arm in the time domain. Results Angle–versus-angle relationships. The relationship between upper arm vs. lower arm angle and torso vs. upper arm angle for a far forward reaching movement in the stable (solid lines) and vibration conditions (dotted lines) are illustrated in Figure 2. Fig 2A compares the control condition with a 4 Hz lateral vibration while Fig 2B compares the control condition with a 6 Hz vertical vibration. It appears that under vibration exposure the reduced upper arm extension is compensated by an increase in torso flexion. This effect is seen in the last phase of the movements (encircled areas). In addition, the lower arm extension is delayed under 6 Hz vertical vibration (Fig 2B left panel). Time of joint motion onset. The timing relationship between torso, upper arm and lower arm is largely a function of the target to be reached. Examples of delays between body links are illustrated in Figure 3. The control condition is compared to a 6 Hz vertical vibration for three

Jan 6, 2006

By John S. McCartney, Ismaail Ghaaowd

This study presents the results from centrifuge modeling experiments performed to understand the effects of temperature changes on the pullout capacity of scale-model energy piles embedded in normally- consolidated clay layers. The energy pile under investigation is an aluminum cylinder instrumented with internal strain gages and thermocouples, and its temperature is controlled using an internal electrical resistance heater. The energy pile was first heated until the temperature, pore water pressure, and volume change of the surrounding clay layer stabilized, after which the energy pile was cooled. Pullout tests were then performed on the energy piles heated to different temperatures to understand the effects of the preceding heating-cooling cycle on the pullout capacity. In addition, the clay undrained shear strength profiles were measured using push-pull T-bar penetration tests before and after heating. INTRODUCTION When a heat source such as an energy pile is embedded in a saturated clay layer, temperature changes will result in changes in pore water pressure as the rate of heating is typically faster than the rate of pore water drainage (Campanella & Mitchell 1968, Abuel-Naga et al. 2007, and Ghaaowd et al. 2016). Although all soils will initially expand during undrained heating, normally consolidated soils typically exhibit permanent contraction while overconsolidated soils exhibit expansion after sustained heating when drainage is allowed (Vega & McCartney 2015). Subsequent drained cooling typically leads to elastic contraction. Thermally- induced changes in pore water pressure and volume during transient heating and cooling may affect the soil- structure interaction of energy piles, as well as the ultimate capacity of the energy pile. This study is relevant as relatively few studies have been performed on the behavior of energy piles in soft clay (e.g., Ng et al. 2014), with most focused on the behavior of drilled shaft energy piles in stiff rock or overconsolidated clays (e.g., Laloui et al. 2006; Murphy et al. 2015; Wang et al. 2015; McCartney and Murphy 2017).

Jan 1, 2018

By Gerald Stolarczyk

The Radio Imaging Method (RIM) is based upon electromagnetic (EM) wave propagation in the coal seam waveguide. The radio geophysics theory underlying RIM was developed by Hill of the national Institute for Science and Technology (Hill, 1984). Recent advances in radio geophysics instrumentation have increased operating range in the coal seam waveguide to almost 3,000 feet. This paper describes RIM longwall panel reconnaissance scans and tomography results.

Jan 1, 2006

Hydrometallurgy has a history of over 300 years with copper leachingand cementation having been exploited in Spain as far back as 1670.Since then, hydrometallurgy has become an integral, expanding branch of extractive metallurgy. This paper reviews major developments in hydrometallurgy over thepast century with emphasis on recent commercial developments and their impact on the mining and metals industries. The environment in which these developments has taken place has changed significantly in the past two decades and many factors, often non-technical, have emerged as constraints on the operation of existing plants and the expansion and development of new hydrometallurgical operations. Finally, a number of possible new developments that will shape thefuture of hydrometallurgy in the new millennium are discussed,particularly as they relate to the assimilation of technical advances in other areas of science and engineering. The impact of current trends in the education of hydrometallurgists and the provision of research in support of these future developments is also addressed.

Jan 1, 2001

By M. J. Senk, A. Lusin

Federal regulations require that methane liberated at the face be diluted and removed to maintain methane gas concentrations below 1 pct (CFR) in working places and intake air courses. The delivery and distribution of intake air at the face is critical for effective face ventilation. Improved techniques for monitoring the movement of intake air to the face are needed. A test system was developed and a computer program written for using a three-axis ultrasonic anemometer to measure airflow in a simulated mine entry. Results of tests conducted in the entry showed that the anemometer is a valuable tool for measuring mine airflow conditions near the face.

By Jo Katembo, Patty Bwando, Patient K. Ilunga, Yves Kalubi, Emman Mulaj

"From 2010, the dolomite concentration in the ore fed to the Usoke Plant increased, resulting in a significant increase in the gangue acid consumption and Mg build-up in the circuit, and thus an increase in the Mg:Co ratio in the plant solutions. This affected the Co salt grade, reagent consumption, filtration and drying capacity. Efforts were put in place to minimise the acid consumption and consequently the Mg dissolution (such as use of pre-diluted acid), but the benefits were negligible. Various efforts were then initiated by CHEMAF to improve the cobalt plant productivity (quantity and quality) by process changes, dual precipitation and seed recycle. The positive results obtained in the laboratory were successfully implemented on plant scale. A change in the process from a single-stage precipitation with Na2CO3 to dual precipitation using a combination of Na2CO3 and NaOH improved the capacity of the Co salt filtration by 30%, the Co salt grade and drying capacity by 50%, and reduced the reagent consumption by 22%. Implementation of precipitate seed recycle increased Mg rejection by 40%, improving the Co salt grade from 20% to 25%, and yielding a further drop in reagent consumption. INTRODUCTION CHEMAF (Chemicals of Africa), a mining company located in Katanga province of Democratic Republic of Congo (DRC), has been operating since 2003. The Usoke hydrometallurgical plant treats the Cu-Co ore from the Etoile pit to produce Cu cathodes via a leach–solvent-extraction (SX)–electrowinning (EW) circuit and Co salt via a selective precipitation process. The plant was designed to treat run-of-mine (ROM) ore grading 6% Cu and 1.5% Co; but, in 2008, due to a decline in grades, a heavy-media separation (HMS) plant was commissioned to enrich the low-grade ROM (3.5% Cu and 0.5% Co) up to 8% Cu and 1% Co before feeding it to the hydrometallurgical plant. Co is recovered by precipitation of Fe and Al with limestone, followed by Co precipitation as Co carbonate using sodium carbonate (Na2CO3), also known as soda ash. 85% of the final effluent from cobalt precipitation is recycled back to the plant as process water; the balance, 15%, is bled out and compensated for by fresh water top up to control and retard the rate of sodium sulfate build up."

Jan 1, 2015