We continue to consider the topic of natural gas pricing and reduction and its impact on heat treaters. Much of the discussion in this month’s article initially appears to deal with process quality or consistency. But understand, process consistency and energy savings are inextricably linked.
This Technical Tuesday column appeared in Heat Treat Today’s December 2021 Medical and Energy print edition. John Clarke is the technical director at Helios Electric Corporation and has written about combustion related topics throughout 2021 for Heat Treat Today.
In February 2022, we will continue this series. Please forward any questions or suggestions to our editor Karen@heattreattoday.com.
Technical Director
Helios Electric Corporation
Source: Helios Electrical Corporation
No matter what method we pursue to save natural gas, it is safe to assume it will require some investment — time and/or materials. Furthermore, we want a payback from this investment. To calculate the payback, we need to estimate the cost of the project as well as the value of the natural gas saved. We can generally nail down the cost of a project by obtaining quotes for materials and labor, but it is more difficult to know what the future cost of natural gas will be; and without knowing the savings, the payback is at best an educated guess.
As we have discussed in previous articles, demand for North American natural gas is increasing for electrical power generation as well as liquified natural gas (LNG) export to areas in the world with limited supplies. These are steady, predictable demands and less susceptible to seasonal variations in temperature. Less heating demand during warmer winters is generally offset by greater electrical power generating demands during warmer summers.
Let us revisit recent trends in the cost of natural gas. The graph below depicts the spot price for 22 consecutive trading days ending November 2, 2021.
Beware of the displaced origin on the graph below — it makes the fluctuations in the spot price appear greater than they are, but it is done to indicate a range of prices — generally around $5.50/mmBTU. (Once again, neither the author nor Heat Treat Today presents the opinion of future prices for any purpose other than to further our discussions of energy saving project paybacks.)
Last month, we posed three questions:
- How do I know when the material I am heating is at the desired temperature?
- Do I have excessive factors of safety built into my process to compensate for not knowing the temperature at the core of the part being heated?
- How much fuel can I save with a shorter cycle?
Much of the discussion in this month’s article initially appears to deal with process quality or consistency. But understand, process consistency and energy savings are inextricably linked.
What temperature is my furnace or oven?
You walk up to the controls and read 1650°F. Is that the temperature of your oven? The answer is a definite “maybe” because the temperature displayed on a single loop temperature controller is simply the reflection of the small voltage generated by one thermocouple. This is obvious, or else we wouldn’t need to run temperature surveys. But the question is — do we have to live with this shortcoming? The answer to this question is a definite “no”! Modern control instrumentation makes it easy to use many thermocouples to sense the temperature of the furnace throughout the chamber. Then take the mean of these values to calculate the temperature and use this average value for the input to our temperature control loop. By comparing the readings of temperatures at various points in the furnace chamber, we can sense if all the work being heated is near to the desired setpoint.
No furnace load is perfect — there is always some non-uniformity of mass or surface area. With multiple sensing points, the more massive and slower to heat portion of the load will influence the nearest thermocouple. The furnace control can be designed to hold until the coolest thermocouple in the chamber reaches some minimum temperature. Perhaps this is now the trigger for a soak timer.
In addition to measuring multiple chamber temperatures and inferring the actual temperature of the work, the proportional integral derivative, or PID, temperature control algorithm provides a good deal of insight as to how close the work is to the desired furnace temperature. All PID controllers or programmed functions provide an output value. For our discussions, we will assume the output is between 0-100%. This output is used to control the heating element(s) of burners’ input levels. The advantage of the PID loop is that it calculates the required value more rapidly than a conventional on/off control — providing us the near steady values for our furnace temperatures.
Let’s imagine we adjust the temperature setpoint of our empty furnace to 1650°F. We will allow it to come to temperature and wait an hour until it is soaked out, so that the refractory and internal components are at some steady state temperature. The PID loop will settle to some average value; we will assume this value is 35%, which represents the holding consumption of the furnace. The heat entering the furnace is in equilibrium with the heat being lost through the refractory, up the flue, around the door, etc.
Now we load the furnace with 4000 pounds of thick steel parts, where the mass/surface area ratio is very high. The furnace thermocouple(s) will reach 1650°F in one hour; but, if we look at the PID loop output, it will take time for it to fall to 35%. The time between the indicated 1650°F and the output falling to 35% is a period when the work continues to absorb heat and conduct it to its core. When the output stabilizes at 35%, we know the work is soaked out at temperature — in other words, the surface and core of the parts are at the furnace setpoint temperature.
Do I have excessive factors of safety built into my process to compensate for not knowing the temperature at the core of the part being heated?
With added insight into the actual temperature of the work being heated, excessive soak times can be reduced without risk. It also allows for the running of light and heavy loads with the same program.
How much fuel can I save with a shorter cycle?
Building on the same hypothetical; assume the input to this furnace is 4,000,000 BTU/Hr and 1,000 hours are saved per year — the savings will be roughly 4,000,000 BTU/Hr x 0.35 (holding consumption) x $5.50/mmBTU x 1,000 Hours per year, or $7,700/year. Now, perform this modification on four furnaces. Add to this savings the increased confidence that the work is at temperature before the soak period is initiated, better consistency for varying part loading, and I think we can agree — we have a project. The only question is, will we cash the check?
About the Author:
John Clarke, with over 30 years in the heat processing area, is currently the technical director of Helios Corporation. John’s work includes system efficiency analysis, burner design as well as burner management systems. John was a former president of the Industrial Heating Equipment Association and vice president at Maxon Corporation.
