Recycled Glass Energy Research Kiln

The energy research kiln in operation 

This project is sponsored by the California Department of Conservation and managed by the Center for Environmental Economic Development

As part of this project we are testing the energy implications of using recycled glass as a raw material in brick manufacturing.  Over the years there have been a number of studies of the use of glass as a raw material to save energy in ceramics kiln processes.  One of the most recent can be downloaded at
One of the reasons these studies have not resulted in widespread use is that all of the previous work focused on using very finely graded glass as an additive to other ceramic raw materials.  The disadvantages of using very finely graded glass include:
  Uniformly crushing glass to fines is relatively expensive,
  Large quantities of uniformly crushed fine glass are not currently available, and
  Very fine glass hurts the workability of clay mixes.

In contrast, this study will test coarser glass, crushed to 12 mesh and finer, as a substitute for grog in brick manufacturing.  12 mesh recycled container glass is readily available as a raw material in fiberglass manufacturing.  For forming bricks, the coarse glass reduces the amount of water needed and accelerates drying.  This work will especially investigate the firing implications for energy use.

A special kiln was built for this study.  The kiln needs to be very fast and accurate.  It is always a challenge to separate the energy issues associated with the kiln itself from the firing issues associated with the material in the kiln.

For good control and monitoring, an electric kiln with quartz halogen heating elements was designed and built.  Quartz halogen elements have virtually zero thermal mass and radiate instantaneously at an equivalent black body temperature approaching 3000 degrees F, peaking in the near infra-red range.  They therefore do not present the complicating factors of slow heat-up and transition periods between conductive and radiation heat transfer that are associated with resistance elements.

Quartz Kiln Construction

Inside of kiln showing quartz support tubes

 The inside area of the kiln is 16 inches square.  The kiln is insulated with 2-inch thick Type M Board made by Thermal Ceramics Corp., contained by an expanded metal frame. 

The quartz halogen bulbs are enclosed in transparent quartz tubes to reduce the exposure of the bulbs to organic burn-off and to cool the bulb ends.  Both of these strategies should help lengthen element life.  Clear quartz tubes are suspended in the middle of kiln, between the upper and lower elements.  This eliminates the need for kiln furniture and reduces he thermal mass of non-fired ware in the kiln to virtually zero.

Control is provided by a Fuji Proportional-Integral-Derivative (PID) controller, with a 4-20 milliamp signal to an Omega phase-angle Silicon Controlled Rectifier (SCR).  Patching into the 4-20 milliamp control signal is an Onset “Hobo” data acquisitor.  Sampling rate is set at one point per second.  Data is transferred to a spreadsheet, where a correlation is made between milliamp output and kiln watt input (unfortunately, SCR input signal to output power is non-linear).  Total kiln wattage is from eight-1600 watt 240 volt quartz bulbs, totaling 12,800 watts at 100 percent output.  Milliamp control signal is limited to 75 percent control output (95 percent power output) to assure long life for the elements.

On initial checkout, the elements proved to be so responsive that it has been difficult to optimize the PID parameters for steady control.  The main control strategy at this point is proportional.  Integral and derivative contributions to control are minor.

Quartz Kiln Usage

The chart below demonstrates the speed of the kiln, as it goes from 1100 degrees F to 1800F in three minutes:

 Left scale is degrees F, bottom scale is minutes

The energy analysis will be performed in several ways.  Samples will be fired containing 50 percent coarse glass as the grog.  After firing, the samples will be tested for absorption.  Absorption will be used as the initial test for brick efficacy.  The firing profile using the minimum amount of energy to make a brick with absorption less than 5 percent will be developed through repeated testing.  Bricks fired using this minimum profile will then be tested for flexure strength as the final test for brick functionality.

Energy will be calculated for kiln consumption at the minimum profile.  Then the kiln will be run empty at the same profile.  The difference will be the amount of energy consumed by the brick during firing.  

Parallel tests will be run using conventional brick grog in place of the glass.  The energy required to make a brick using grog will be derived in the same manner.

Energy Analysis

Since the data derived from the Onset data recorder is in time/rate format, it may be possible to derive the difference between endoergic and isoergic energy requirements.  The chart below illustrates the concept.  The kiln was fired empty, then using the same profile (up to 1800∞F) with ten pounds of glass in molds.  The energy was monitored during both firings.  The upper curve represents the energy requirement during the firing with glass in the kiln.  The lower curve is the empty kiln.  It can be seen how the kiln needed more energy to perform the glass firing.

When the two curves converge, that means that the temperature of the load and the kiln are equal.  The load is “done.”

The violet line is energy consumption with load, blue line is energy consumption empty.  Left scale is watts, bottom scale is minutes. 

The specific heat of the glass with molds is
    .22 btu/∞F-pound
Heating glass is an isoergic process, that is, vs. endoergic or exoergic processes, no net energy is either consumed or given off.  For ten pounds of glass, the amount of energy expected to heat the glass from 70 to 1832∞F is:
    .22 btu/∞F-pound x 10 pounds x (1832-70)∞F =  btu = 3876 btu

During actual tests, the empty kiln consumed 6257 watt-hours empty and 7455 watt-hours full.  The difference the load in the kiln made was:
    7455 – 6257 = 1198 watt-hours x  3.4 btu/watt-hour = 4073 btu.
The predicted value is within 5 percent of the experimental value.  This is a remarkable correlation.  Even this error may be reduced with a better correlation between the 4-20 milliamp signal and the actual power output from the SCR.

A strategy like this will be used to derive the optimal firing profile for ceramic mixes including recycled glass as a raw material.  The final report will be distributed to manufacturers of ceramic products, with the hope that they will consider introducing recycled glass as a raw material.

next page