ABSTRACT
Industrial processes use rotating equipment (e.g.; pump, fan, blower, centrifugal compressor, positive displacement compressor) and pipe (or duct) to move fluid from point A to B, with many processes using electric motors as the prime mover.
Most of the systems in the industry are over-designed to meet a peak load demand which might occur over a small fraction of the time or to satisfy a higher pressure demanded by a much smaller user in the same process. The system over-design will result in a selection of larger but inefficient rotating equipment and electric motor system. Some of the typical design examples that result in higher energy usage are given below.
A careful life cycle cost and economic evaluation must be undertaken to ensure that the process audit, reengineering and equipment selections are not impacting the industrial process goals, but result in a least optimal cost over the life of the project.
The paper will define, discuss, and present various "process systems" in chemical, hydrocarbon and pulp & paper industries. It will discuss the interactive impact of the changes in the mechanical system configuration and the changes in the process variables to better redesign the system and reduce the cost of operation. It will also present a check list of energy conservation measures ("ECM") or opportunities. Such ECMs will be related to hydraulics, system components, process modifications, and system efficiency. Two or three case studies will be presented focusing on various conservation measures that improve electrical operating efficiency of a distillation column system. An incremental cost and payback analysis will be presented to assist the investment in process optimization and energy savings' measures.
A typical fluid flow system consists of a mechanical system and a process system. A Process Flow Diagram (PFD) defines the heat and material balance for that fluid flow system, and a Piping and Instrumentation Diagram (P&ID) defines the mechanical and design configuration of the same system. For example, a mechanical system of a compressed air system may consist of various end-users of compressed air, delivery or distribution system (piping, valves, fittings, controls, etc.), compressor (centrifugal or positive displacement), its associated equipment (aftercooler, receiver, dryer, filter, etc.), and a driver (electric motor).
The process system will consist of air pressure and temperature at end-users, quantity and quality of air requirement, physical properties (molecular weight, polytropic exponent), ambient air temperature and air pressure, moisture in the suction air, etc. The energy consumption is impacted by the design of mechanical and process systems as well as the end-user load profile and hours of operation. It is very important to understand the technical equations used in the calculations of flow rate at various standard temperature and pressure (STP) options, inlet air density, outlet air density, compressor capacity, adiabatic or polytropic head, discharge temperature, break horsepower, and compression ratio. A further discussion of a recent article on compressed air can be found in Energy Engineering, (Volume 95, No. 6, 1998) and therefore, another case study from a petrochemical industry will be presented in the next section.
A design engineering firm normally evaluates the three operating scenarios at best; namely, the normal flow, the maximum flow, and the turndown flow. The pipe, fittings, and pumps are selected using the maximum flow rate, but control valves are sized to operate under the all three flow conditions. This may result in control valve pressure drop at the normal flow rate to be as much as 30-40% of the system pressure drop or 10-15 psi minimum. The electric motor is usually a next larger size than the one estimated at maximum flow rate condition. In addition, a single pump is generally designed to serve many users. The sum of all load is the pump flow capacity. The total dynamic head of the pump is, however, controlled by one user which either requires a higher delivery pressure or is far away (such as tank farm) with an undersized pipe line and/or the user has a high static head requirement. In any case, that flow loop becomes the controlling head for the pump selection. A control valve in other circuits or loops will consume the additional pressure drop. An engineer can start preparing an off-line model of the system and develop a baseline or current operation model once the process boundaries are drawn and set points (related to pressure, flow, level or even fluid composition) are established. A spread sheet or a computer software program can assist in the baseline model development. The friction losses due to pipe, fittings, and valves are calculated. An estimation or field measurement of pressure drop in overhead condenser, heat exchanger, filter, etc. is made. Once the system model is verified, then various "what if" case studies are conducted. The goal is to develop an optimal model of a fluid flow system that meets the production targets as well as minimizes the energy usage under various operating scenarios. Most chemical plants and refineries operate 24 hours a day, 7 days a week, year around with minimum planned downtime. The following section outlines a case study problem statement for fluid flow systems analysis and its analysis of energy savings' options.
A process engineer or an energy analyst is evaluating a distillation
column that is separating binary mixture of benzene and toluene.
The column, with an overhead pressure of 5 psig, produces a distillate
containing 99 benzene and a bottoms product containing 95 weight
percentage toluene. A Product/Reflux pump (P-101) is sized for
843 gpm at 280 ft of liquid head. Reflux flow is 2.54 times the
distillate rate or 605 gpm. Benzene product flow to off site tanks
is 200 gpm. A 38 gpm product flow is directed to another process
unit. All process parameters and elevations are shown in the Figure
1.
Figure 1. Base Case Operation & Design of
Benzene/Toluene Distillation Column With a Single Pump
The pump P101 is designed for three delivery duties: Reflux flow, Product flow, and Process flow. The pump suction and discharge piping with fittings and valves are sized and laid out that followed generally acceptable good engineering practice. The flow system has three control valves. The reflux back to the tower is controlled by a flow control valve (FCV-1), product discharge flow to the storage tank is controlled by a level controller (LCV-1) from the accumulator and flow to another process unit is set by a hand control valve (HCV-1). There are other control strategies available but they are out of scope for this article. The system also has other equipment such as heat exchangers, air cooler, and tanks. The pump P101 is a Goulds model 3x4-13 LTX, 3196 with a 75 hp electric motor. The certified pump curve information is available in the maintenance engineering file and is used in the analysis to plot systems curve and characteristic curves at different operating scenarios.
The following above grade physical elevations and the delivery
pressures (Table 1) were noted on the P&IDs and verified in
the field.
Table 1. Base Case Elevation and Delivery Pressure
| Elevation above Grade, Ft. | Delivery Pressure, psia | |
| Reflux line to top of column | ||
| Pump centerline | ||
| Accumulator liquid level | ||
| Storage tank normal level | ||
| Process unit |
The goal is to evaluate and optimize the process design, and to reduce the energy usage. The analysis will create more than one case study. Each case study will have one or more changes in Process System as well as Mechanical System. Interactive effect of more than one ECM will have to be factored in the energy usage. For this analysis, we will assume that steady state flow requirements will not be changed. We will primarily focus on splitting the pump P-101 duty into two parallel pumps, one dedicated to reflux flow only and another pump for product and process discharge flow. We will also evaluate implementation of a variable frequency drive (VFD) replacing flow control valve, and re-plot the pump characteristic curve, system curve and estimate energy savings. Appropriate changes to the control valve pressure drop will be made to meet the process design requirements. Other ECMs can be larger pipe sizes, rerouting of lines to reduce length, but those changes can be expensive and may not result in good implementable projects. A check list of possible ECMs is presented in a later section.
It is important to model the base case accurately to establish which flow circuit (end-user) out of three flow circuits controls the pump selection. The pump flow is sum of all three end-user flow requirement, but the pump head is determined from the worst case end-user demand, which, in this case is the reflux flow circuit from the Pump 101 back to the top of column. The flow control valve pressure drop at normal flow is 30 psi. The two other flow circuits will have the same pump discharge pressure, but the control valves (LCV-1 and HCV-1) will be sized to consume that additional head. At normal flow rates, the level control valve in the flow circuit to tank will have a pressure drop of 64 psi and the hand control valve in the flow circuit to another process unit will have a pressure drop of 94 psi.
The following Table 2 (Pump Calculation Sheet) provides the "controlling"
base case operation and Table 3 lists the major hydraulic parameters
of three base case flow circuits (reflux flow, storage tank flow,
and process flow).
Table 2. Pump Calculation Sheet For the Base Case Operation
Table 3. Hydraulic Parameters of Base Case Single Pump Service
It is important to understand and model the existing operation using either a spread sheet or a computer software program specifically designed for the fluid flow systems analysis. This should be followed by the analysis of one or many processes and mechanical component variables to evaluate the impact on power usage by these changes. A brief check list of energy conservation measures is given below that can be applied to pump, fan, blower, and compressor systems. We will evaluate some of the measures in the following sections.
We will investigate three ECMs in detail and evaluate their impact
on energy savings. Please note that there will be many other measures
(singular or combination) that can be investigated. The goal,
however, must be to pick the most cost effective measures for
analysis. These measures will have a greater chance for implementation.
The following measures are analyzed further.
The following Figure 2 shows how the benzene/toluene distillation system will be served by two pumps. All other mechanical and process parameters will remain the same as the base case.
Figure 2. Two Pumps in Parallel
The Table 4 on the next page summarizes the key variables of the base case and ECM 1 case, and estimates a 13.6% energy saving. The energy saving is due to a reduced pressure drop in LCV-1 and HCV-1 control valves. The alternate pump P-101 is a Goulds model 3196 MTX, size 3x4-10 process pump. It has a slightly better mechanical efficiency (72.8%) than the base case pump. The new product pump (P-103) has a 63% pump efficiency at the normal flow rate of 238 gpm at 135 ft of head. In all case studies, a 90% electrical motor efficiency is used. It should be noted that upgrading a standard efficiency motor to premium efficiency motor would only gain 2-4% in operating efficiency. However, upgrading and selecting a right pump would significantly increase the operating efficiency by as much as 5 to 10%. An energy engineer should focus on fluid flow systems components (e.g., motor, pump, VFD, pipe, control valves, etc.) that could improve system efficiency dramatically while keeping the cost down.
The controlling hydraulic case for the pump P-103 is product flow to the storage tank via two coolers in the series. As you will notice, an additional electrical energy saving can be achieved by reconfiguring two coolers from series to parallel operation (eliminates 10 psi pressure loss) and increasing the line size to the storage tank (reduces~6 psi).
Table 4. Hydraulic Parameters of Two Pumps in Parallel (ECM
1)
It is assumed that the control system response time permits a replacement of the current flow control valve by a variable speed drive. The VFD will control the reflux flow that is either set by column overhead temperature, pressure or composition. During the normal operation, the system will see the reduction in system pressure profile equivalent to control valve pressure drop (30 psi). Figure 3 plots system curve and pumps curves at normal speed as well as at variable speed.
It should be noted that during turndown and reduced reflux flow
rates, the savings will be even greater with a VFD. For example,
at reduced reflux ratio of 2.0 (versus base case of 2.65), the
reflux flow will decrease to 476 gpm. This will result in reduced
discharge pipe frictional losses to 43.31 feet from 69.9 feet
(P flow2). The flow control valve will pinch close
to absorb the additional pipe pressure loss and higher head generated
by the pump curve at the lower flow. In this specific case, the
system gains discharge line friction loss of 26.59 feet (=69.9-43.31),
the exchanger frictional loss drops from 10 psi to 6.19 psi, and
the orifice pressure drop is now 3.1 psi from base case 5 psi.
The pump head (without a VFD) would be 99.56 psi or 286.15 feet
of liquid. The flow control valve, FCV-1, now has a pressure drop
of 52 psi, an increase of 22 psi over the base case operation.
A VFD, with a 96% efficiency, would save 15.62 kW or 49% savings
over the case if no VFD was used but only the reflux flow was
reduced from 605 gpm to 476 gpm. This case is not further presented
due to space limitations.
Figure 3. Pump Curves and System Curves (ECM 2) Using C-MAX
software
The base case pipe friction losses from pump P 101 to the top
of the column is 69.9 feet. The discharge pipe consists of two
4" diameter schedule 40 pipe segments with a 120 feet straight
length & number of fittings. The friction losses are calculated
using "non-compressible flow Benoulli's equation (Reference:
Crane TP 410 manual & C-MAX software). At normal benzene flow
of 605 gpm, the 4" pipe has a fluid velocity of 15 feet per
second. Increasing this pipe to 5" schedule 40 pipe would
decrease the velocity to 9.70 fps and reduce the frictional losses
to 19.8 ft, saving additional energy. The Table 5 tabulates the
hydraulic and energy savings parameters for ECM 2 and ECM 3.
Table 5. Hydraulic and Energy Savings Parameters for ECM 2
and ECM 3
The Table 6 summarizes the energy savings realized by conducting
a systems analysis of the fluid flow system. An incremental cost
and payback analysis in the Table 7 clearly indicate the attractiveness
of these ECMs. An electricity cost of 8 cents/kWh and a demand
charge of $ 5/kW/month are used in the calculations.
Table 6. Summary of Energy Savings
Table 7. Cost Estimate and Payback Analysis
The continuous reengineering and process optimization are necessary as industrial outputs change to meet the market demand. A company that implements an aggressive plant energy management and demand side management program will reap the benefit of operating cost saving, which always flows to the bottom of the income statement. A penny saved in NOT equal to a penny earned!! For commercial businesses, a penny saved is equal to 20 pennies earned (revenue) because the net income approximates to 5% of the revenue. The investment payback for efficiency improvement and savings projects could be justified on the basis of revenue that need not be earned to stay in business rather than on the basis of return on investment (ROI) of 2-3 years for the new capital projects. The bottom line to evaluate energy savings opportunities that should result in improved deployment of current and future capital investment.