Considerations for Flow Computer Selection

By Irvin Schwartzenburg
Fisher Controls International, Inc.

Overview

In today's highly competitive electronic gas flow measurement market, purchase price is often used as the dominant selection criteria. Initial purchase price alone is not a good indication of long term ownership cost. Factors affecting this cost include product functionality, installation requirements, application versatility, reliability, maintainability, ease of use, and supplier support and services among others. This paper outlines these factors and shows how they can impact a system's true cost.

The primary and most obvious purpose of all automated electronic gas measurement systems is to measure gas flow accurately and reliably. In addition to performing the actual gas measurement they must provide local data storage, audit trails and means to collect this information.

Companies are actively embracing the future by looking at automation as not only a means to eliminate the paper chart but as the only way to survive in the future. Often these systems start out as basic electronic gas measurement systems. Enhancements in the area of communications, monitoring and control can be added to these basic systems to help realize the full value of the system. Once the substantial investment is made to put an infrastructure in place to retrieve and process field information it then becomes a natural extension to offer these services to other operators in the region. Daily production, tank levels, well control, alarm status and other reports are provided as a daily service allowing their use of manpower in the most efficient way possible.

Basic Components of Electronic Flow Measurement

Industry recognizes some general definitions for common components used in electronic gas measurement systems. These components consist of the physical meter (primary device), process transmitters (secondary devices) and the flow computer (tertiary device). Common metering devices are the turbine meter and orifice meter. Transmitters provide signals correlating to the gas quantities being measured. Static pressure, differential pressure and temperature are three common process variables that are measured by transmitters. The flow computer takes the information produced by the metering device and transmitters, along with user-entered inputs, and produces a record of the gas quantities that have passed through the measurement point.

In addition to the metering devices, thought should be given to area classification, site security and power as each one of these can greatly affect ownership cost.

Metering Devices

To achieve the goal of accurate and reliable gas measurement, steps should be taken to eliminate as many potential sources of error from the flow equation as possible. The primary metering device is where the measurement begins and should be given careful consideration when trying to maximize EFM performance. There are several different types of primary metering devices, but the orifice meter is perhaps the most use often used method. One of the most common problems with orifice metering is the bent orifice plate which is typically caused by sudden pressure changes and slugging. A large deformation can produce errors exceeding 19% of true quantities. Other common problems include reverse installation of the plate, poor eccentricity, sharpness (or lack of) of the orifice and its cleanliness, the orifice bevel condition (scratches, nicks), the meter tube condition and its installation. Even on a small run averaging 40 MCF per day at an estimated gas price of $2.00 per MCF, a 15% error could cost you over $4,000 a year in gas that is unaccounted for.

Transmitters

Choosing the correct transmitters for your application can significantly reduce the long term ownership cost of your system by improving its accuracy and reducing maintenance costs. Several options for transmitters are available to companies pursuing electronic gas measurement. They range from built-in transducers to external "smart transmitters". Thought should be given to the advantages of each with respect to accuracy, stability, interchangability, temperature effects, static pressure effects and repeatability.

Today, smart transmitters offer the best value for the money. Since smart transmitters talk digitally to the flow computer's processor, they eliminate two places where errors can be introduced into the flow equations. In a conventional analog transmitter the signal is converted from the analog process variable (such as pressure) into a digital value by the device's analog-to-digital (A/D) converter. The transmitter then conditions this value and converts it back to an analog value (typically 4 to 20 mA) by means of its digital-to-analog (D/A) converter. This analog signal is then converted by the flow computer's A/D converter into a digital value for use by the flow equations.

With smart transmitters, the process variable is converted into a digital value by the transmitter's A/D converter. All compensations to this value are made by the factory characterization information residing in the smart transmitter. The final process value remains in its digital form and is read directly by the flow computer's processor. Therefore, if the transmitter reads a process value of 81.250 Inches of H2O, this is the exact value used by the flow equation. Accuracy can be greatly compromised if this value is sent through a conventional transmitter's D/A converter and then through the flow computer's A/D converter. Smart devices usually have an accuracy approaching three times that of typical analog accuracy without taking into account the potential error introduced by the process variable's voltage or current signal and the A/D conversion.

A smart transmitter retains the extensive factory calibration and characterization within its non-volatile EEPROM. This allows smart transmitters to be field replaceable or to be moved from one flow computer to another without losing factory characterization. Conventional transmitters usually have the sensor characterized to the flow computer at the factory with the characterization residing within the flow computer. Thus, if damage occurs to the sensor, the complete unit must be sent back to the factory for re-characterization.

Stability and repeatability can also affect long term cost. A smart transmitter can have a stability specification of four times that of a conventional analog transmitter. When a stability of 0.2% upper range limit (URL) for 6 months is compared to a stability of 0.1% for 12 months, it is easy to see that an additional source of error input to the flow equations is eliminated. Additional savings are realized by eliminating the high maintenance cost associated with keeping conventional transmitters in calibration. Operations overhead is reduced because your skilled manpower can be used in ways more productive that constantly re-calibrating transmitters.

Multi-variable transmitters and transducers are now commonly available in flow computers. They measure both differential and static pressure and may even provide a temperature value all in a single device. Not all multi-variable transmitters are smart devices, so be sure you understand what is being provided if important to your application. The major advantage of the multi-variable transmitter is cost. A single multi-variable transmitter can replace three conventional transmitters at less than half the cost.

Calibration is also a cost issue. High accuracy transmitters dictate the use of high accuracy verification and calibration equipment. For example, API Chapter 21, Section 1.8.6 gives guidelines for calibration and verification equipment. It states "The minimum uncertainty requirement for calibration/verification equipment shall be a factor of two better than the specified uncertainty of the transmitter...". It also states that in a practical sense, an accuracy of greater than +/- 0.05% is normally not required. Therefore, if an investment is made in high quality sensors, verification and calibration equipment should be accurate enough to correctly prove the process readings.

Calibration equipment used to provide +/- 0.05% accuracy for differential pressure readings are usually dead weight testers. Devices that can calibrate static pressure transmitters to this accuracy are not as common and can be quite expensive when pressure ranges exceed 300 PSI. For these installations, many companies will use an NIST certified device for verification in the field and have the re-calibration, when necessary, done in a controlled environment. This is possible when smart transmitters are used since they maintain their own characterization and interchangeability is not a problem. You cannot calibrate a 0.075% transmitter with a calibration device that is 0.1% accurate any more than you can record seconds with an hour glass.

When you take into consideration accuracy, repeatability, maintainability and initial cost of smart multi-variable transmitters, it is easy to see the long-term cost advantage they provide. Accuracy improvements over conventional transmitters alone can eliminate the potential for error in excess of $5,500 / year on a 10MM gas run. Combine this with the elimination of error introduced from the 4-to-20 mA loop or 1 to 5 VDC signal conversion, the lower maintenance cost, and the lower initial cost of a multi-variable transmitter, and you have an exceptional value.

Flow Computers

Prior to the advent of electronic flow computers, circular chart recorders were the mainstay of flow measurement. A paper chart recorded differential gas pressure across an orifice plate. These charts were gathered and sent in to the company's gas accounting group either on a weekly or monthly basis for integration. The integration process was slow and "off-chart" gas flow could not be accounted for. Records were kept in the field office showing verification and calibration information. Gas samples were taken quarterly and analyzed, and the analysis information was forwarded to the gas accounting department for use in chart integration. The gas accounting group did the rest.

With flow computers, the responsibility to enter and maintain information used by the flow computer's flow calculations and the maintenance of the subsequent audit trail information is shifted from the office to the field. Other responsibilities being moved to the field are data editing and recalculation. The office still handles the data processing, accounting and archiving.

Issues pertaining to the physical installation, calculation options and the choice of data averaging techniques must be understood and addressed before choosing a flow computer. This is to ensure compatibility with internal requirements and existing systems as well as to reduce the installation time. The flow computer should have the means to easily transfer its configuration to other like devices once these questions have been answered. Field programming and flexibility are also a must. While no two wells are alike, runs within a gas field may be identical with respect to physical installation but can vary in gas composition, control and monitoring needs and contractual obligations. As you move between gas fields these variances are even greater.

It is therefore prudent to choose a flow computer that is flexible in configuration and expandable in I/O to address all possible requirements of the company. Important factors that should be considered when choosing a flow computer are:

• AGA configuration options. Older contracts based on the 1985 AGA equations are still in place while new contracts are moving toward the 1992 standards. Does the flow computer allow the calculation options to be chosen by the customer and set up on a per run basis? Are there any hard coded parameters? If so, ensure that they meet your contractual obligations.

• Historical Averaging techniques. API Chapter 21 defines four different methods of historical averaging techniques. They are flow dependent linear, flow dependent formulaic, flow weighted linear and flow weighted formulaic. Will the flow computer support these techniques and are they user selectable? Once again, if it does not, what is the default and is this the best method for your application?

• Future expansion capability. How adaptable is the flow computer to different situations? Projects that start out as basic gas measurement systems may evolve into monitor and control schemes requiring odorant injection, flow and nomination control, well "blow down" control, plunger control, condensate handling, treater monitoring and even compressor control to name a few. It is to your advantage to consider your company's automation vision when choosing a flow computer so as to select the unit best to suited your long term needs. Substantial investments in hardware, software, training and installation may be lost if dedicated single function units have to be replaced to meet future requirements.

• Operating system storage. Flow computers that utilize newer technologies such as "Flash" ROM storage instead of EPROM allow their firmware to be upgraded in the field or over remote communications links by a simple file download. Compare this capability to removing the unit from service and replacing or "burning PROMs". When multiplied by several hundred units it is easy to see the savings that can be realized by not having to visit each site.

• Configuration security. The ability to have specific site configurations stored in EEPROM provides the best possible means to protect against configuration loss in the event of a memory initialization, also know as a "cold start". A unit that can be re-started by reloading its configuration from EEPROM can provide significant savings in time and cost for large installations.

• Historical database format and configuration. Because historical requirements change from company to company, you'll want a unit that allows you to define which points are archived.

• Communication port availability. Units that provide multiple communication ports that are both hardware and software configurable provide tremendous flexibility by allowing connections to several types of communicating devices. For example, such a unit can communicate to a host both through a radio or leased line, while at the same time communicate to an intelligent device such as a gas chromatograph.

• Communication Protocols. Closed communication protocols can lead to a single-path approach to automation. A flow computer using a closed protocol will severely limit your choices when integrating the device into a new automation system. It can also be difficult and expensive to integrate the flow computer into a existing system if its protocol is protected by the manufacturer.

• Local keypad access. A keypad that is attached to the flow computer can simplify access for information editing and calibration which in turn can increase operator efficiency. Common activities such as changing orifice plate constants, verifying data values, calibration and setpoint changes can be done faster through a local keypad than through a hand-held device.

Other Issues Affecting True Cost

While gas measurement is the primary function of a flow computer, other issues should be taken into account when determining the value of one flow computer over another.

Approvals and Classifications

All flow computers should carry a certification as to the Classification, Division and Group they are approved for service in. Common agencies that provide approvals are Underwriters Laboratory, (UL) and Canadian Standards Association (CSA). When CSA approval is used, make sure that the NRTL/C mark appears next to the rating. Provisions have been made where CSA and UL can now test to each others specifications and obtain approval for both organizations. Thought should be given to the future of the installation and how it fits into the overall automation vision of the company.

For example, most flow computers must be derated for Class I Division 1 use once communications are added. If this is considered in the original installation, the costs associated with moving the unit when features are added can be avoided.

Site Security

API Chapter 21 details the type of security required for data access and data integrity. It was once said by a wise man "locks were made for honest people". If physical security is a concern, intrusion switches used in conjunction with a report-by-exception communication scheme are suggested so that security violations can be logged and reported immediately by the flow computer. Not all flow computers can support this type of security.

Power Supply

For a system to be reliable it must have a dependable power source. The common choices are between AC and solar systems. A properly sized power system must take into account both power consumption and autonomy. Power consumption determines the load on the power supply and should take into account constant and peak loads. Autonomy, also referred to as battery backup time, is the amount of time the flow computer can operate on battery power without any external charging. The incorrect sizing of batteries and/or solar arrays for a required load is perhaps the biggest factor for premature battery failure, resulting in higher overall system life-cycle cost.

The most common battery type used with solar and line power backup systems is the sealed lead-acid gel battery. This battery has three characteristics that determine its life expectancy. First, when 80% of it rated capacity has been discharged, its voltage has probably dropped to an unacceptable level, usually less than 11 VDC. This is referred to as an 80% depth of discharge. Second, up to 70% of available battery capacity can be lost in sustained ambient temperatures of -30 degrees C (-22 degrees F). Third, frequent cycling to a depth of discharge in excess of 50% can reduce battery life by as much as a factor of three.

The number of times that a battery can be cycled, depending on its depth of discharge per cycle, can range from 200 to 1500 cycles. On solar-powered systems, it is critical to ensure that battery capacity and panel size are sufficient to account for ambient temperature swings and typical winter conditions to prevent excessive daily discharge. Failure to do so means you will be replacing batteries prematurely which directly relates to higher lifetime cost. In a poorly designed application, if a high depth of discharge occurs on a daily basis, batteries can fail in a little as 200 cycles. This means you may be replacing several thousand dollars worth of batteries that would still be alive if they were properly sized to the application. Depending on cost per battery, a 500 site installation could have a yearly battery replacement cost in excess of $15,000 in batteries alone. This doesn't even take into account the time and labor involved in replacing them.

Communications

The most frequent enhancement to basic EFM, and the most beneficial, is the addition of communications to a host system. Communications allows measured data to be collected and parameters, such as gas analysis, to be remotely entered. For example, site data is made available to operators so that they can better plan their daily tasks. Alarms can be reported on sensitive installations so that corrective action can be taken and well control and nominations can also be handled.

Common methods of communicating to flow computers are through dial-up modems, leased-lines (Bell 202), radio (licensed and unlicensed frequencies), microwave and even satellite. Combinations of these can be used as well. Each method has its advantages and disadvantages. Frequency of polling and speed should be balanced with cost. Site requirements may dictate communications to be close to real-time or as seldom as once a month.

With radio systems, it is especially important that enough time and money be allocated to design and implement the most dependable system possible. Generally, communication speed and availability relate directly to cost. Some EFMs support a communications scheme that can substantially reduce power system requirements, resulting in a battery and panel size reduction of over 50%. Even a modest reduction of the power system of $100 dollars per site on a 100 site project can result in savings of $10,000 on the initial project cost. A poorly designed system can waste thousands of dollars in troubleshooting and maintenance cost.

Supplier Support

This can be a make or break issue in the successful implementation of an EFM system. The company or companies with which you are dealing should have a reputation for good product quality and support and be financially stable. You'll be making a sizable investment in hardware, software, installation and training. You should question the longevity of the company and its commitment to you in addition to its commitment to future technology development. Frequent or recent ownership changes should concern you. The worst thing that could happen is for a company to change its long term focus during your project.

Local support is an area where many reputable companies fall short. Questions you should ask are what type of local support is available from the supplier and what is the satisfaction level of other customers with the support provided to them? What is the typical response time that the local support company can provide in an emergency situation? If the company cannot provide evidence showing a high level of local support, you may want to consider another supplier.

Conclusion

In conclusion, up-front cost alone is not the best measure of true flow computer value. Value must be measured against such factors as accuracy, reliability, user acceptance, functionality, ease of use, application of open standards and supplier support. It is only when all of these factors are taken into account that you can determine the true cost of owning your EFM system.