May-2016

21st century reciprocating compressors for downstream applications

Technology, like time, marches on. Think about the advancements in technology that have taken place in the last 50 years. Advancements have taken us from punch cards to medical computers that can be swallowed.

Greg Phillippi
Ariel Corporation

Article Summary

Telecommunication has gone from rotary dial to cell phones. Most advancements led to smaller, less expensive means to do the same thing. It begs the question, why do some industry standards and the people who write them, refuse to acknowledge proven technologies that are as reliable but less expensive than those used 50 years ago? Take the reciprocating compressor industry as an example.

For nearly 100 years reciprocating compressors used in downstream oil and gas industry (process) applications have utilised long strokes and low rotating speeds. Virtually all have been block mounted and equipped with compressor cylinder liners and provision for cooling supported since 1964 by American Petroleum Institute Standard 618 “Reciprocating Compressors for Petroleum, Chemical, and Gas Industry Services”.

Today alternate designs are available offering the industry equal reliability at lower capital cost. Several manufacturers offer designs with shorter strokes, in the 76 to 229 mm (3 to 9 inch) range, with rotating speeds varying from 600 to 1000 rpm. These modern designs are packaged into complete compression system modules contributing to reduced capital cost through reduced installation time and therefore cost. Many of these modern designs omit liners and provision for cooling as required by API Standard 618 further reducing cost.

Very early reciprocating compressors had strokes in the range of 914 mm (36 inch) or longer, had rotating speeds that today would be considered to be very low, in the range of 100 rpm, and were derived from steam engine technology. Over time the rotating and average piston speed increased as the sealing technology for piston rod packing and piston rings improved. For example, very early piston rod packing was called a “stuffing box” because it was a cylindrical cavity surrounding the piston rod filled (stuffed) with coiled rope to create a seal. This “stuffing box” was derived from steam engine technology. Over time the packing became much more sophisticated developing into the segmented packing ring sets made of various metallic and non-metallic materials so common today.

There is end-user interest to increase rotating and piston speed because increases in both lead to reduced capital cost. Increases in rotating and piston speed reduce the physical size of the machine resulting in less mass and therefore less cost. Today, the vast majority of downstream process reciprocating compressors are driven by electric motors, which have their cost decrease as their rotating speed increases. So an increase in rotating speed decreases the cost of both of the compressor and the driver.

Block Mount versus Packaged
The physical size of the typical long stroke (305 to 508 mm, 12 to 20 inch) low speed (250 to 500 rpm) compressor, as shown in Figures 1 and 2, lends itself to being stick-built at site (block mounted) rather than packaged into a module in a fabrication facility. Most of the reciprocating compressors that exist today in refineries and petrochemical facilities are block mounted meaning bare compressors are built in a factory, shipped to the installation site assembled or disassembled (depending on size and shipping restrictions), and installed on a large concrete foundation (the “block”, see Figure 1). After which all the supporting systems, such as pulsation bottles, separators, process and utility piping, lubrication systems, driver, coupling, instrumentation and control system, are installed. Long stroke low speed compressors lend themselves to this manner of installation because they are typically very large and heavy.

Modern short stroke medium speed compressors lend themselves to being packaged (as shown in Figure 3) as they are smaller and lighter for the same capacity. A compressor package is a complete gas compression system module having the compressor with its driver, coupling, all process gas and utility piping, lubrication systems, and all the instrumentation and control systems mounted on a structural steel skid. This skid serves as a platform on which to mount all the previously mentioned equipment, but also, in many instances, the compressor’s foundation.
Lined versus Unlined

API Standard 618 requires compressor cylinders to have liners. Fundamentally, liners are included only for commercial reasons. A liner is not a part required for a compressor cylinder to be able to compress gas. The reasons a liner may be used include:
• A liner can be a lower cost replaceable wear element in a compressor cylinder assembly where a bare replacement cylinder body might be very expensive with a long lead time relative to the cost and lead time of a replacement liner.

• A liner can be made of a suitable wear material, such as grey iron, when the cylinder body is made of an unsuitable wear material. An example of such an unsuitable material is ASTM A395 “Standard Specification for Ferritic Ductile Iron Pressure-Retaining Castings for Use at Elevated Temperatures”, the material API 618 requires be used for cast ductile iron cylinder bodies. ASTM A395 happens to be a very poor material for use in an application that subjects the material to rubbing wear as is the case in a cylinder bore where the piston and piston rings, or the wearbands and piston rings, are rubbing against it. Consideration must be given to protect A395 ductile iron material and a liner is only one way to accomplish that.

• Utilisation of a liner allows the cylinder bore diameter to be changed rather easily.
One of the major reasons why liners are not utilised in short stroke cylinders is the liner’s effect on the cylinder’s capacity capability. This is a technical issue that leads to significant commercial harm. The following chart, Figure 4, will help to explain.

A liner adds fixed clearance which reduces capacity. This capacity reduction can be substantial in shorter stroke cylinders. The addition of the liner also reduces piston displacement, which when combined with the effect of the additional fixed clearance, results in a significant reduction in a cylinder’s capacity capability. The Figure 4 chart takes both into account. For example consider a 300 mm liner bore diameter (11.8 inch, “300” on the horizontal axis). This is actually a 325 mm (12.8 inch) diameter bore cylinder with a 12.5 mm (0.5 inch) thick liner installed, so the capacity capability has been reduced from that of a 325 mm bore cylinder to that of a 300 mm. Also, the addition of the liner has increased the fixed clearance volume, which reduces the volumetric efficiency further reducing the capacity capability. For the 300 mm liner bore example this capacity capability reduction is on the order of 35 percent for 76 mm (3 inch) stroke, and is still significant for a 457 mm (18 inch) stroke at about a 17 percent reduction. The chart assumes a compression ratio of 2.5 and a gas adiabatic exponent of 1.4. Assumed liner thickness is 9.5 mm (0.375 inch) up to 254 mm liner bore (10 inch) and 12.5 mm (0.500 inch) 254 mm bore (10 inch) and larger following API Standard 618 guidelines.

For a given required compressor capacity, the addition of a liner requires a larger compressor by the percentage shown in the chart. Again referring to the 300 mm example, a 76 mm stroke compressor would have to be about 35 percent larger if equipped with a liner. So an end-user is buying a compressor about 35 percent larger just to have it equipped with lined cylinders.

While a liner provides one method of protecting A395 ductile iron from wear, other methods exist. Another possibility is to harden the unlined cylinder bore to improve the wear characteristics. One proven hardening method uses the ion-nitride heat treat process. A full explanation of the process is beyond the scope of this article but it results in a hardness at the surface of A395 ductile iron of approximately 55 Rockwell C and provides substantial case depth with a hardness of approximately 30 Rockwell C at a depth of 0.15 mm (0.006 inch).