March 2019 Newsletter

Writing a Good RFQ

Understand your requirements The first thing is to understand your requirements.  Now this does not mean you have to perform the design or have all the detail.  But if your request is clearly laid-out, the systems integrator does not need to include a lot of contingency costs in the quote for the unknown or changing requirements.  Less risk almost always means less cost for you.  Unknows equal higher risk for the bidder, and higher costs for you. Usually one or two meetings with all the stakeholders at your facility will produce a thorough understanding.   And sometimes after this is complete, you will find that your project will take on a whole new complexity – sometime less complex – sometimes more.  Maybe you will find that there is no problem at all or the solution is right in front of you. List the milestones You should put a timeframe to the project.  For example, Phase X needs complete by Date A and Phase Y needs complete by Date B.  Rush projects also mean more risks; and more costs to you.  When possible, get your RFQs out before the situation becomes an emergency. Clarify Implementation Constraints If your project has no specific implementation window or does not require after-hours or weekend work, make that clear as it can save money. Apples-Apples Ask the same things of all your vendors who will be asked to quote.  If all bidders are qualified and bidding on the same requirements, it is much easier to select the lowest bidder. Be honest about your requests If your request is a ‘fishing trip’ or to plan your future year budget, make that clear. (Picture 2) It takes a lot of effort to produce a quote for a custom-designed system, so wasting your vendor’s time (and the time of their customers) can be costly down the road when you have an actual project requirement.  Being clear about your intentions gives the vendor options concerning their participation.  Reputable suppliers want to provide goods/services at a fair price.  They are looking for win-win situations.  Writing a good RFQ makes everyone successful.

 

Temperature Controller Calibration Checks

Most manufacturers have a requirement to check that their temperature controller inputs are correct to some standard.  The requirement maybe based on a customer requirement, an industry certification, or a best practice.  The input devices are typically thermocouples, but also maybe RTDs.  If you are looking to implement a calibration check procedure, there are two common methods to consider.  Both will require a calibration instrument (aka calibrator) with an optional NIST traceable certificate. Quickest but Least Functional Method Using the calibrator in the measure mode, the measuring leads are touched to the controller input device terminals to gain a reading.  The reading is then compared to the controller reading and a +/- tolerance is applied to see if the controller is within specification.  For example, a calibration procedure has a tolerance of +/- 10 degrees F.  The temperature controller is reading 1700⁰F and the calibrator is reading 1688⁰F.  The controller is considered out-of-spec and requires a corrective action (either replacement or adjustment).  The advantage to this method is that the thermocouple does not need to be disconnected from the controller, so the equipment can stay in service during the check.  The disadvantage could be that the controller is only checked at one point (i.e. the temperature that the process is currently at) and no assumption of measurement linearity over the range of temperature can be assumed.  This may not matter if your process stays at the same temperature all the time.  More Functional and Longer Method The Thermocouple must be disconnected from the controller.  The calibrator leads are then connected to the controller.  The calibrator is then used to simulate two or more temperature signals to the controller input, usually a low and hi (and sometimes a mid-range).  The readings on the controller are compared with the signal values from the calibrator.  Then a +/- tolerance is used to determine if the controller is within specification.  The advantage is that linearity can be determined.  Another advantage is equipment that is down and at room temperature, is given a more thorough check.  The main disadvantage of course, is that the equipment must be down in most cases and wires must be disconnected/reconnected when using this method.