When OEM's performance data differ from the user's test results

March 22, 2020
TMI Staff & Contributors

Turbomachinery Blog features postings from experts in all areas of turbomachinery, such as: gas turbines, machine diagnostics, materials, repairs, and aftermarket parts, and encourages users to participate, with reader engagement and interaction as its primary purpose.

Consistent, accurate methods to calculate flange to flange centrifugal compressor polytropic performance are required to evaluate a machine’s thermodynamic acceptance guarantees. These guarantees are typically listed in project purchase documents by reference to an industry accepted code, but only mention general details of implementation.

This article contains excerpts from the paper, "Centrifugal compressor performance making enlightened analysis decisions," by Fred Evans and Spencer Huble of Chiyoda International Corp. at the 2017 Turbomachinery & Pump Symposia.

An original equipment manufacturer (OEM) uses internally developed standard testing procedures and methods and provides test results to a client based upon them. These vendor procedures are rigorously controlled for consistency across a product line. When a client develops an independent determination of performance test results that yields answers different from the OEM that may exceed specified tolerances, resolution can be difficult since it may involve technical as well as contractual aspects. By this time it is usually too late in a project’s schedule to make any hardware changes without additional costs and long delays.

The time to address performance determination methods is before a contract is signed, not after test stand results are brought into question. It may seem less important during the period of contract negotiations, but writing details into a clause in the contract can help avoid problems or help deal with them efficiently later. Having a clear understanding of why various performance calculations can yield different results from the same input data can go a long way in resolving test stand issues.

The origin of differing thermodynamic performance calculation results can usually be traced to seemingly minor variations in developing polytropic exponents, real gas correction factors, gas properties determined from various equations of state, and occasionally, the very definition of efficiency. Compressor testing can take many forms to which these calculation methods can be applied. The most common testing scenario occurs at the manufacturer’s facility just prior to hardware shipment when the compressor must be shown to meet design requirements.

Due to facility limitations and safety precautions, performance testing does not typically duplicate field design conditions but is based upon aerodynamic similitude employing inert gases, typically at lower than design pressures and power. Consequently, performance calculation methods that are sufficient for processing test stand measured data may not be as applicable for field measured data.

The Power Test Code (ASME PTC 10, 1997) provides a framework for centrifugal compressor testing and prescribes allowable testing tolerances. Maintaining test results within these tolerances is intended to allow conversion for comparison to project specified design conditions via similitude. The code is meant to assure an accurate test result but does not directly address technical acceptability of a compressor to meet contractual requirements. In other words, adherence to the code helps perform a good test that is representative of a particular machine’s capability.

Further evaluation of achieving contractual guarantees is not the intent of the power test code as those are governed by contractual documents and typically tolerances found in API Standard 617 (2014). Within the code framework, the processing of measured test stand data can be accomplished by the choice of three methods. The most commonly applied is based upon a classic technical paper (Schultz, 1962) that derived real gas thermodynamic relations for the compression process using suction and discharge end point fluid states and a correction factor derived from isentropic considerations.

Two methods for determining polytropic exponents were defined in Schultz’s paper along with an example case. The Schultz methodology was adapted and incorporated into the ASME PTC 10 code in 1965 and reaffirmed in 1997 and has been employed by many users when purchasing compressors. However, exact rules for implementation within the test code are not provided for all aspects, leaving the door open for a certain amount of interpretation such as mentioned in code Section 5.2.

Assumptions employed by Schultz to arrive at a solution can be shown to cause inaccuracies beyond acceptable limits in some applications. Over the ensuing years since first introduced, several technical papers have offered comments, refinements, improvements and/or alternatives to the Schultz methods. Also during this time, the determination of real gas properties via thermodynamic equations of state has vastly expanded in quantity and quality, as well as ease of access and use.

When the Schultz method was first embraced by the PTC 10 code, availability of computing power was a mere shadow of what it is today. Performance calculation methods no longer need be hampered by simplifying techniques since robust numerical manipulation requires only inexpensive yet powerful resources that are readily available to any analyst.