February 19, 2001
Installing a bulk delivery system in your welding shop is perhaps the best way to save money
Editor's Note: This is the first of a two-part article that discusses shielding gas efficiency. The second part also appears on thefabricator.com.
Although argon is the most commonly used shielding gas for solid-wire electrode applications, many managers and welders understand little about this inert gas and its costs. Most welders are told argon makes a slick weld, leaves no flux, is expensive, and can cause suffocation in enclosed environments. Most also are told to conserve this gas by cutting it off on their gas tungsten arc welding (GTAW) torches during breaks, lunch, and at the end of a shift.
Argon generally is quantified by the term centicubic feet (CCF). One CCF of argon is equal to 100 cubic feet of argon gas - - the amount it takes to fill a 4-foot by 7-3/4-inch box.
If the cost of argon is $2.50 per CCF, it costs only $2.50 to fill the box with argon at sea level. Filling the box with argon at 45 cubic feet per hour (CFH) would take 2 hours and 13 minutes. Because a flow rate of 30 to 45 CFH is adequate for most gas metal arc welding (GMAW) applications, one might say that $2.50 is not a lot of money for 2 hours and 13 minutes of continuous welding. In a perfect world, this would equate to about $9.00 of gas cost per weldstation per eight-hour shift at 100 percent arc time.
Assuming most manual weldstations' actual arc times are from 15 to 30 percent, not 100 percent, the cost is reduced to $1.35 to $2.70 per eight-hour shift. This does not sound like much, but these figures represent a perfect environment with just one weldstation and one work shift during one workday. When other circumstances are considered, those 135 to 270 pennies can increase rapidly.
For example, 300 weldstations operating with 100 percent efficiency at 30 percent arc time, three shifts per day, 312 days per year will use $758,160 worth of argon per year. However, most plants operate at less than 50 percent efficiency, which, in the previous example, means spending more than $1.5 million annually on welding gas.
Some plants mistakenly assume that because consumption does not vary much from year to year and production levels remain constant, efficiency has been maintained. After all, gas costs are built into the cost of the finished product. This can be a very expensive assumption.
Most welding engineers seem to agree that argon gas (or a mixture) is the only choice. However, even with its many advantages, if used unwisely, argon rapidly can become a double-edged sword.
Buying argon and other gases in large quantities and in a liquid state reduces the unit cost per CCF. Compared to typical 336-cubic-foot steel cylinders, gas for a bulk system can be purchased for about half as much per CCF. Additionally, the amount of labor saved from not having to manipulate cylinders and production downtime to exchange cylinders will help pay for a bulk system's installation cost fairly quickly.
If a plant uses a bulk system, it probably can reduce its gas consumption rate. Many plants can reduce consumption by 50 to 80 percent or more in a single year by instituting a series of conservative measures; there is no single solution. However, the upside is that even when multiple measures are taken, none are particularly expensive compared to the benefits. The return on investment generally can be measured in days, not in months or years.
Often, the first step is making employees aware of the company's annual consumption costs. This data then can be compared to other consumables (primarily wire) that are directly proportionate to gas use. This can be accomplished by determining an average wire speed (in GMAW applications) for a plant or plant area. This wire speed can be used to calculate the amount of shielding gas required to burn 1 pound of wire at a given gas flow rate at the torch tip. For example, assume a 0.045-inch-diameter mild steel wire is used at an average wire speed of 300 inches per minute (IPM) and a shielding gas flow rate of 35 CFH. If about 2,210 inches of 0.045-inch-diameter mild steel wire is required to equal 1 pound, the following applies:
(300) (60) / 2,210 = 8.14 pounds of mild steel wire per hour
Because the gas flow rate is 35 CFH, the gas-to-wire ratio is 35 divided by 8.14, or 4.29 CFH of gas to 1 pound of wire (4.29-1) Thus, if a plant's annual wire consumption is 500,000 pounds, the annual shielding gas consumption should be about 2,149,500 cubic feet.
Remember, this scenario represents a 100 percent consumption efficiency and does not necessarily represent a real-world environment. Several factors can affect this method of comparison such as:
Most plants that consistently manufacture the same mild steel products are fairly consistent with wire speeds and sizes. These plants generally can verify consumption data, and the only nonwelding consumption to verify is waste. This leaves item 2 (large flow variations) as the most prevalent factor in making this type of comparison.
In these types of plants, most welding engineers and other experts in the field agree that a 10-1 ratio or less is acceptable even though this is more than twice the ratio given in the previous example (4.29-1). Some plants, according to their data, have ratios as high as 55-to-1 and as low as 7-to-1 initially. Many plants can reduce their ratios from the 18-to-1 to 30-to-1 range, down to the 9-to-1 to 14-to-1 range by addressing some seemingly minor issues, discussed in a later section.
If properly designed, installed, and maintained, a bulk system can offer many cost and productivity advantages compared to conventional cylinders.
Bulk systems should be engineered, designed, and constructed with several things in mind. First, they should be of a closed-loop design that can handle all present and future flow requirements with the minimal amount of system pressure and pressure drop across the entire system. The system should be designed to minimize hose requirements and provide maximum protection from external impacts, yet still remain easily accessible for inspection, modifications, and repairs. Construction materials should vary, depending on each area of the system's application.
Copper pipe with silver-phosphorus joints serves well for most applications. However, steel pipe with both screw joints and welded joints also works well in some systems, depending on the severity of the environment. Where welded joints are used, the socket weld type generally is better because most maintenance departments do not have qualified open butt pipe welders to make modifications and repairs. All forms of PVC pipe should be avoided if possible.
System drops should originate from the top of the header and run down to a small manifold with an isolation ball valve located directly above it. These isolation valves should be installed so that any type of gravitational force exerted on the valve handle will close the valve. Each drop should be supplied with the appropriate number of outlets and an accompanying isolation valve. All unused outlet valves should be closed, as well as plugged.
These drops should be accessible but not necessarily convenient. Employees often use them to hang coats, grinding shields, hats, and hoses on. This practice can lead to an inadvertent release of shielding gas that can go undetected for a long time. For this and other reasons, drop headers or manifolds normally should be of a larger size pipe than that of the drop supply line and, when possible, constructed of brass, steel, or other hard material that is resistant to deformation from abuse.
All fittings should be threaded, 300#-rated (or higher) forged steel. Drop manifolds should be securely mounted and equipped with a minimum 6-inch drip leg that can be removed for blowdown purposes and to provide a means for future header modifications without shutting down the entire system. Y fittings on drop header outlets also should be avoided.
Often in older plants, the existing shielding gas piping and distribution system has outgrown the goals of its original design. This primarily is because of plant expansion, internal modifications, changes in the shielding gas used, etc.
Accurate documentation of all piping is necessary to conduct flow analysis on any system. This documentation should contain the location and types of all valves, pipe size changes, pressure regulators, and all corresponding dimensional measurements. In most cases, major modifications are not required to bring a system up to par.
The type of devices that should be used to regulate flow to each weldstation is an area of considerable debate. Plants using adjustable flow devices or rotormeters must ensure that the required flow delivery rates for the weld procedure stay within a reasonable range. This is even more critical in nonbulk systems. Inspection of hundreds of plants that use the rotormeters for flow regulation reveals less than 20 percent are set to the proper flow delivery rate.
Typically, this type of meter will deliver up to 450 CFH at or close to the full open position (depending on model and system pressure). Just because the indicator reaches the top of the sight glass does not mean the flow rate stops increasing if the valve is opened more.
For the example given earlier regarding the perfect-world shielding gas cost of $2.70 (at 45 CFH) per eight-hour shift, the cost now is $11.25 per shift. Even for a one-shift operation, the annual shielding gas cost increases from $842.40 to $3,510.00 per weldstation. Annually, the plant's argon consumption balloons from $252,720 to more than $1 million. By setting an adjustable flowmeter at or near its maximum open position, a plant's shielding gas consumption can increase tenfold.
The reasons that these rotormeters are often open fully vary. In the summer months, welding personnel often have more ventilation or fans blowing directly on them and thus increase the flow rate to maintain their shielding gas purge. When colder weather arrives, the fans disappear but the flowmeter setting does not change.
Some welders think that "if a little bit is good, then a lot is even better." This is not necessarily true. Depending on torch-to-work angles, this high-velocity jet of shielding gas can actually induce atmospheric contamination of the weld puddle and create more weld contamination problems than it solves. Also, it is wasteful.
Rotormeters always should be mounted on the hard piping drops located at each weldstation. The length of the typical 1/4-inch hose that runs to the wire feeder should be considered. The hose generally offers a lot of friction coefficient because of its internal composition. In addition, the hose usually is routed up, down, and all around, which restricts the gas flow. Unless it follows a straight line, the hose should not exceed 25 to 30 feet.
If rotormeters are mounted at or near the wire feeders, the mounting location should be rigid, upright, and out of harm's way. Rotormeters are not very impact-resistant and when mounted this way, often become the source of leaks and can cause gas contamination.
Another risk associated with rotor flowmeters mounted at the wire feeder is leak severity. Hoses can develop leaks upstream from the meter with a combined flow rate that is much higher than a rotormeter will allow to pass, even in the fully open position. If a flowmeter is mounted at the drop and the hose develops a leak, the meter restricts the flow, reducing gas delivery to the torch, which would become apparent to the operator.
If the flowmeter is mounted on the wire feeder, the leak is constantly subjected to line pressure, with the flow limited only by the size of the leak opening and the operating pressure. This configuration gives no indication of upstream flow and usually results in the operator compensating by increasing the flow at the meter.
This scenario also introduces atmospheric contamination into the system over a period of time. As the number of occurrences increase, the original design parameters are overtaxed, resulting in a greater pressure drop across the system. This, in turn, usually results in increased overall system pressure to offset the higher pressure drop. This action increases the across-the-system pressure drop even further and magnifies the severity of all the system leaks and other losses.
Regardless of its mounting location, the rotormeter-type flowmeter should be kept in readable condition and subjected to its proper calibrated pressure. It is not unusual to find flowmeters calibrated for 20 pounds per square inch gauge (PSIG) installed in systems operating at 60 PSIG. This can result in delivery flows as much as 15 to 18 percent higher than the delivery flow indicated by the flowmeter scale. This practice, on a plantwide basis, can be very expensive.
The initial flow surge encountered at the torch tip when the solenoid valve is activated on the wire feeder also should be considered. This flow surge typically is associated with flow devices mounted at the system drop and/or when a large amount of hose or other piping is used between the flow metering device and the wire feeder's solenoid valve. This larger internal volume is subjected to main system pressure when the wire feeder is not in use.
When the wire feeder is activated, the accumulated pressure bleeds off rapidly from the torch tip and gradually decreases to the amount set by the flow device. If the flow device is mounted closer to the torch tip (at the wire feeder), this internal volume is minimized, which decreases the time required for the flow to reach the set flow rate. This, in turn, decreases the amount of gas wasted by momentary excess flow or overflow. In some robotic and other high-cycle-rate applications (tack welds, etc.), this overflow can be substantial.
For example, assume a flow device is located on a drop off of the main system and 15 feet of 1/4-inch-inner- diameter (ID) hose is connected to the wire feeder. When the wire feeder is not in use, pressure in the hose rapidly builds to the system pressure of, say, 30 PSIG. When the wire feeder is activated, the hose pressure drops to almost nothing (depending on solenoid valve porting). This wastes approximately 0.01 cubic foot of shielding gas as excess gas until stable flow is established. Based on the example plant mentioned earlier, this waste costs about 3/100 of one penny per occurrence.
Now assume that the hose length is increased from 25 to 75 feet. The value of this overflow now equals 14/100 of one penny each time each wire feeder is activated. Again, this type of loss doesn't seem like much per occurrence, but when it is multiplied by the number of wire feeders plantwide and the cycle rate or number of times each wire feeder is activated daily, it can become very significant, very quickly.
Using round numbers for simplicity, assume that each wire feeder in the example plant is activated once every minute. This equals 8 (hours) 60 (minutes) 300 (welders), or 144,000 activations per shift plantwide. At a cost of 14/100 of a penny per occurrence, and with three shifts operating, the total cost is about $540 per day and more than $168,000 annually.
A single shielding gas loss appears to be insignificant, but when it occurs over the long term, its cost can be significant.