Alternative Bath Control Practice

Electroplating facilities analyze process baths to maintain product quality. Analysis is performed with Hull cells, test strips, titrations, meters and other tools. Bath additions are calculated as the product of the bath volume multiplied by the difference between the analytical result and a bath component specification. Materials are then added to the process bath in a way and a time that does not adversely affect product quality. In short, process control for most facilities is the practice to ‘add-on-analysis’.

 

A well run finishing facility will have a documented program for process bath analysis, bath additions, record keeping and preventive maintenance. The components of a process bath are checked on a schedule. Only authorized materials are added at a suitable time.  Both the readings and the additions are recorded in a paper or electronic log. Those baths that exhaust over time are changed on a schedule.

 

Operating experience will show that certain components change more rapidly or are more critical to product quality. To maintain product quality, it is desirable to minimize the difference between the measured values and the specifications for the process bath components. With the practice to ‘add-on-analysis’, this range is minimized by increasing the frequency of the analysis and the subsequent additions. Obviously, frequent analysis increases the cost in labor and materials. 

 

There are statistical tools to evaluate gauges, detect drift and change in a process, to separate out the most critical parameters and to determine interactions among parameters. Just as with the analytical tools, there is a cost in labor. For many facilities, the short lead times and the rapid pace of production presents a challenge to achieving the full benefit from the application of these statistical tools.

 

I suspect many facilities experience quality issues from time to time in spite of their best efforts to maintain the process to specifications with the practice, to ‘add-on-analysis’. There is another way to process control that is based on the concept of ‘mass balance’ and ‘the first derivative’. Your facility should already have the necessary data and you should save labor. The concept presented here leads to a whole host of operational and cost saving benefits.

 

Let us review what factors cause changes to the component concentrations in a process bath. First, solution is removed from the process as work and work carriers exit the bath. Second, certain components become part of the product. Third, some components are changed by the process and are no longer beneficial. Fourth, components are removed to adjust or while treating the process bath. The first, commonly referred to as drag-out, equally affects all components of the process bath.

 

For most captive shops and many job shops, the production is predictable. Each work center is designed to finish certain part types that require finishing in some customary quantity on a day to day or week to week schedule. Those parts that are processed less often also tend to be supplied in lower volumes. This is referred to as the ‘80:20 rule’ which is that 20% of the part numbers will usually occupy about 80% of the production schedule. This is the key assumption. If production continually varies in part type and volume, the following proposed change from the usual process control practice will not work for you.

 

If the production is predictable, then so should be the material consumption. The consumption of a majority of the components of each process bath should be proportional to the amount of work processed. For example, the relationship between production and component consumption is well established for plating bath brighteners. These proprietary additives are fed with auto feed systems based on rectifier amp-hour monitoring.

 

If the consumption is predictable, then the additions can be planned. That statement is the cornerstone to this alternative process control practice. To maintain a process bath in specification, component consumption must be replenished with timely material additions. That is the essence of a ‘mass balance’. However, these material additions do not need to follow after an analysis as is the standard practice. In my view, the analysis of a process bath should be used to evaluate and adjust a planned schedule of material additions that are initially set to the historic rates of consumption.

 

 The title was chosen because this bath control concept is analogous to the operation of cruise control on a vehicle. The cruise control device adjusts the throttle to raise or lower engine power so that the vehicle speed is adjusted toward the set point. Process bath analysis should be used to throttle an established schedule of material additions. And just as cruise control on a vehicle tends to lower fuel consumption, this process control practice should reduce material consumption by avoiding spikes in concentration that cause greater losses from drag-out. 

 

A further justification for this change in procedure is the accuracy of that process bath analysis. Examine your records of analysis and material additions. I suspect that you will find that on occasion a component concentration has increased beyond any material addition; or alternatively, a large addition is made and there is little change. This is a hint that the usual methods for analyzing a process bath in a production setting may have some error from a number of causes. Yet additions are routinely calculated based on the difference between the analysis and the specification.  If the analysis was off, then the calculated material addition could be detrimental to the cost of operation or worse, to the product quality.

 

On the basis that periodic process bath analysis has been performed and that there are records of production, of the analysis and of the subsequent material additions, the first step to implement cruise control is to determine rates of consumption, ‘the first derivative’. Calculate the totals for all the process bath additions that have been made over the past three months or longer. Subtract any additions related to some special circumstance. Divide those totals by the work center production hours or some other measure of production volume for the same time period. The result is your average consumption per hour or per an alternative chosen measure. For instance, you may choose consumption per plating barrel. Since these rates of consumption are determined over some extended time period, any errors in the analysis should accumulate toward a mean. A PC based spreadsheet is an excellent tool to perform these calculations and others that build on this information.

 

Next, schedule the material additions at regular intervals and in a quantity equivalent to the calculated historic consumption rate. The frequency should be set so that the additions are small relative to the original makeup, but large enough to keep the work load reasonable. Automated feed is favored relative to manual additions if the additions must be small and frequent to maintain product quality, such as with plating bath brighteners.

 

For example, the log records that 300 pounds of boric acid was consumed at nickel bath N9 in the last quarter. The boric acid concentration in N9 is determined by titration once each week. The consumption rate from the bath N9 is calculated to be about five pounds per day based on a five day work week. So, schedule a five pound addition of boric acid each day as long as the production pace remains near norms. For this material, add as soon as the production clears at the end of the work day and add near the discharge point of the filter system. If no production was processed, skip the addition. In the usual practice of add-on-analysis, boric acid additions were probably 20 to 30 pounds per week on the day of the lab work or on the last day of the work week.

 

Continue the once a week analysis to confirm the addition rate and adjust as needed if the concentration begins to drift from the target. Trust the plan and reconfirm analysis that deviates from expected results. Over time, the frequency or addition size may change. There may be a need for a ‘spot add’ or a ‘skip’ caused by an unexpected change in production or special circumstance. In the normal add-on-analysis practice, the bath concentration varied between 30 pounds light to a concentration near the target. In the cruise control practice, the bath varies no more than about 5 pounds light. In time, analysis may be less often and changes to the material addition plan will be caused by planned changes in production volume.  

 

Further, those 300 pounds of boric acid consumed in the quarter indicates that about 800 gallons of plating bath was lost to drag-out and maintenance. That assumed the boric acid is added properly and maintained near the target concentration of 6 ounce per gallon. All the other components of the N9 bath were lost in proportion. So the bath should have also required about 16 gallons of wetter over the quarter, with the wetter at a 2% by volume concentration. Schedule an addition of 1 gallon of wetter per week. Continue to check the concentration on the same schedule and make the small adjustment that may be required. Extra wetter may be required because of a carbon treatment, one of the special circumstances.   Again in time, the analysis may be less often.

 

Some of the components of a process bath are difficult to analyze or data regarding consumption is scarce. Each component exists in the process bath in some desirable proportion to the others. When historic consumption rates are not available, an alternative is to add such components in proportion to another component that has good historic data and is consumed in the same way. Technical bulletins for process bath may provide such guidance as in, ‘for each part of A, add 3 parts of B’, because the concentration of B is difficult to determine. To continue the example above, a small amount of brightener could be added in proportion to the wetter to compensate for brightener lost to drag-out, separate from what is consumed in the plating process and fed with the feeder system.

 

And a reminder, there are process bath components that are not consumed in proportion to production. Some materials are added to a process bath because of operating conditions. A good example is ammonium hydroxide for a brass bath, which may be added because the bath has not been worked and needs to be replenished for loss by vaporization. It is still beneficial to track the consumption of such additives as will be explained later.

 

One means to reduce the work load is to schedule the additions of a particular component to all points of use, which has the further benefit of reducing the opportunity for cross contamination. For example, complete the additions of nickel sulfate across all the nickel baths before proceeding on to the wetter additions. This minimizes the set up time to complete the material addition plan. Since material additions are scheduled in a quantity appropriate to each process, there is the opportunity to balance the work load through the week. Further, a process bath type N that is acidic may be serviced on Tuesday and type CN with cyanides on Thursday, with those persons who perform the additions wearing different PPE for those days.

 

If bath type N is scheduled to be serviced on Tuesday, then any analysis of that bath type must be completed that same morning if not the working day prior. So, just as the labor to service the process bath is balanced through the week, so does the lab work. The tendency of this cruise control process control practice is that key personnel responsible for lab work and bath maintenance fall into a routine and work effectively and efficiently. Supervision has the same milestones week to week to benchmark task completion. 

 

The schedule for planned material additions will evolve. The same kind of thinking is appropriate for adjusting addition rates as is used to evaluate an x-bar r chart for an individual process or product specification. The desirable pattern is a scatter of concentration measurements slightly above and below the specification. This scatter pattern is a result of an actual variation in the bath concentration and an analysis error factor.  A series of concentration measurement all above might indicate skipping one addition to center the component concentration or reducing the rate of additions if there is a climbing pattern. As the rate of scheduled material additions dial-in to deliver consistent concentration measurements, the frequency of analysis can be reduced and the cost of the process control will lessen.

 

And now, what are the other benefits to this process control practice? The consumption rate information may be used in a number of ways to manage the business and reduce cost.   In time, your facility will develop a plan for material additions for each component of each process bath on each work center. Begin with less critical materials like the boric acid of the example and work toward proprietary additives. Those consumption rates were determined on the premise that the process baths were kept in good working order to produce product to specifications. The plan for material additions should produce even more consistent results.

 

With the consumption rate information and a schedule of material additions implemented, a sudden change in bath concentration is a red flag. The issue may be operational, equipment or personnel. There is the obvious concern of an unreported loss of the process bath. Were the scheduled material additions completed as planned? Has there been a change in operators on a manual barrel line that increased drag-out losses? Was there a change in barrels to smaller perforations? Has a leak developed in a heat exchanger? With the practice to add-on-analysis, even a significant change in bath concentration is often seen as normal due to the duration of time since the last check.

 

With the consumption rate information, it is now possible to optimize the material inventory. For example the record may show that boric acid is used in seven plating baths. The total consumed per quarter is 1500 pounds. If no extraordinary work is scheduled on any of those seven plating baths, than a stock of ten to fifteen 50 pound bags of boric acid should be in inventory for the coming month with a delivery schedule of ten per month for replenishment.  Determination of inventory continues material by material. Calculate the expected total consumption for a period equivalent to the time between deliveries. Set the stock level to an amount equal to the consumption plus a safety factor.  Share the data with your supplier.   

 

With the consumption rate information and the data that it is based on, the person responsible for environment reporting has a good foundation from which to work. As described in the previous paragraph, a plan for inventory will develop. There are triggers for the reporting of hazardous material. Comparison to the inventory plan may present opportunities for reducing the reporting burden under the rules, and at least there will be justification for the quantities kept on hand.   As this relates to the release report, the data and derived information indicates where and how hazardous materials were used, which helps in determination of where the hazardous materials went.    

 

With the consumption rate information, material cost and a part production record, it is now possible to determine the material cost for each finish on each work center. The plating material cost could be included in a bill-of-materials in the resource planning system. Add up all the materials consumed in a process bath for a period of time multiplied by the purchase price of each to calculate the material cost over the period. Add up all the individual process bath cost utilized in a particular finishing sequence. If necessary allocate the process bath use across multiple finishing sequences based on relative production volumes. Divide this grand total of material cost by a measure of the production, such as the total area or total weight, finished over the same time period. The result should be a number expressed in dollars, something like $1.50 per square foot to finish from raw steel to semi-bright nickel at work center #9.

 

This derived material cost per finish and per work center has uses in turn. One particular work center may have a significantly lower material cost than another producing the same finish. The work center may have features that reduce or recover drag-out, for instance. In the short term to reduce expenses, maximize production across the more efficient work center. Longer term, the difference in material cost may provide sufficient financial justification for implementing the same drag-out handling features on the other work center. Early in my career, I discovered a close correlation between plating barrel age and material cost as expected. In that facility, the material cost doubled when the barrels reached five years relative to new. The difference justified the purchase of new cylinders for each line after five years of operation.

 

And finally, how does all this help with product quality, which is the whole point. It is the central premise of ‘six sigma ‘that to reduce variation in the process will reduce variation in the product quality. As control of the process improves and delivers a more consistent product quality, it is then possible to improve product quality through tested process change. The problems that require troubleshooting are usually caused by the accumulated affects of variation in the process. Usually returning some key operating parameter or bath component to specification corrects the problem. Every control that reduces variation in the process, such as this cruise control practice on process baths, limits the possible causes of a product defect and accelerates resolution of the problem. That saves cost.

 

In conclusion, understand the purpose of each component of a process bath. Understand how each component is consumed. Plan the material additions to process baths. Use the analytical tools and assessment of product quality to adjust the plan. Minimize variation in process inputs to minimize variation of the process outputs, namely product quality. Strive for continuous improvement. 

 

Louis Krull has a degree in chemical engineering from Northwestern University and has over 25 years experience with metal finishing equipment and process, process control and troubleshooting, regulatory compliance and lean manufacturing. He can be reached at louiskrull@aol.com , or at 860-563-7885.