Ammonia refrigeration systems are typically large energy users and hence good targets for energy efficiency improvements. Common opportunities to reduce energy use and improve energy efficiency include reducing refrigeration load, increasing suction pressure, employing dual suction, decreasing minimum head pressure setpoint, increasing evaporative condenser effectiveness, and reclaiming heat. Estimating savings from these measures is often difficult because of the complexity of the equipment and controls. However, the fundamental energy response of ammonia refrigeration systems to these measures can be estimated using a simplified model. An understanding of these responses is often an important first step to designing or optimizing the performance of actual systems.

Thus, this paper develops a simplified model for simulating a single-stage ammonia refrigeration system and uses the model to estimate how energy use changes in response to the opportunities mentioned above. The results show how system energy use and potential energy savings vary with ambient wet-bulb temperature. The results give designers and operators insight into possible savings opportunities.

INTRODUCTION

Ammonia refrigeration systems are common in commercial and food processing sectors, including grocery, bottling, meat packaging, and brewing. About 7.5% of the total manufacturing energy consumption is used in food processing industry (EIA- 2006 Energy Consumption by Manufacturers). In these facilities, the ammonia refrigeration system is usually the largest energy consumer.

Most ammonia refrigeration systems are customized for a given application. As a result, these systems can be extremely challenging to analyze and model for energy savings. Stoecker (1998) provides theory and detailed engineering calculations for many specific components in ammonia refrigeration systems. Manske et al., (2001) provides a comprehensive analysis and detailed results for optimizing a particular ammonia refrigeration system. Cascade Energy Engineering, Inc. (2010) combines refrigeration theory and field experience to provide a manual of best practices for industrial refrigeration.

This paper describes a simplified model for simulating a single-stage ammonia refrigeration system and uses the model to estimate how energy use changes in response to common opportunities to reduce energy use and improve energy efficiency. It begins by developing a model for a single-stage ammonia refrigeration system, including models for the evaporative condenser and heat reclaim. The model was developed in MATLAB (Mathworks Inc. 2012a) using the Reference Fluid Thermodynamic and Transport Properties Database (NIST 2010) for refrigerant property data. The psychrometric properties were modeled using methods from ASHRAE (2009). The model is used to simulate energy use from a base case system and show how it varies with ambient wet-bulb temperature. The model is also used to simulate energy use of a system with six energy efficiency opportunities in comparison to the base-case system. The paper then demonstrates how annual energy savings could be estimated using bin data from typical weather files. The results show how system energy use and potential energy savings vary with ambient wet-bulb temperature and give designers and operators insight into possible savings opportunities.

MODELING SINGLE-STAGE AMMONIA REFRIGERATION SYSTEMS

Refrigeration Model

A single-stage ammonia refrigeration system consists of five main devices: the evaporator, compressor, evaporative condenser, high pressure receiver, and an expansion valve. A single-stage system may have multiple devices of the same type to meet system requirements. A system diagram for a single-stage ammonia refrigeration system with high and low temperature evaporators with dual compressors is shown in Figure 1a, and a pressure-enthalpy (P-h) diagram for the system is shown in Figure 1b.

The system model uses the following assumptions: a) ammonia leaves the evaporators and enters the compressors as a saturated-vapor, b) expansion through the expansion valve occurs at constant enthalpy, c) pressure drops through the system (with exception of the expansion valve) are negligible, d) heat gains and losses through the piping and high pressure receiver are negligible, and e) the evaporative condenser fans are controlled with variable frequency drives (VFDs). Based on these assumptions, compressor power and evaporative condenser fan and pump power can be calculated by developing a model with nine input values. These nine input values are the cooling loads of the low and high temperature evaporators, the low and high suction temperatures, the isentropic efficiencies of the compressors, the condenser capacity, the ambient wet-bulb temperature and motor efficiencies. The system model is explained below.

At state 1L, ammonia enters the low temperature compressor as a saturated-vapor at the low suction temperature, [T.sub.1L]. These two independent properties are used to fix state 1L and determine enthalpy, [h.sub.1L], entropy, [s.sub.1L], and pressure, [P.sub.1L].

Ammonia leaves the low temperature compressor as a super-heated vapor at state 2L. The isentropic efficiency of the low temperature compressor, [eta][c.sub.L], and the enthalpy of ammonia leaving the compressor under isentropic conditions are required to calculate the properties of ammonia at state 2L. For an isentropic process, the enthalpy of the ammonia leaving the compressors, h[s.sub.2L], is calculated using [P.sub.2L] and [s.sub.1L]. The enthalpy at state 2L is calculated as:

[h.sub.2L] = [h.sub.1L] + (h[s.sub.2L] - [h.sub.1L])/[eta][c.sub.1] (1)

Ammonia enters the high temperature compressor from the high temperature evaporator as a saturated vapor at state 1h. Ammonia properties at state 1h are fixed using the high suction temperature, [T.sub.1h], rather than [T.sub.1L]. Likewise, the enthalpy of ammonia leaving the high temperature compressor, [h.sub.2h], is calculated using the isentropic efficiency of the high temperature compressor, [eta][c.sub.2], the enthalpy determined at state 1h, [h.sub.1h], and the enthalpy of ammonia leaving the compressors under isentropic conditions, h[s.sub.2h] as:

[h.sub.2h] = [h.sub.1h] + (h[s.sub.2h] - [h.sub.1h])/[eta][c.sub.2] (2)

Ammonia leaves the evaporative condenser at state 3 as a saturated liquid at the system head pressure. These two independent properties fix state 3 and determine enthalpy, [h.sub.3], entropy, [s.sub.3], and temperature, [T.sub.3] also referred to as saturated condensing temperature (SCT).

The expansion valve is modeled as a constant enthalpy device. The high pressure receiver is assumed to be adiabatic with negligible pressure drop. Therefore, the enthalpy entering either the high temperature evaporator, [h.sub.4h], or low temperature evaporator, [h.sub.4L], will have the same values as the enthalpy of ammonia leaving the evaporative condenser, [h.sub.3]. Thus, independent properties [T.sub.1h] and [h.sub.3] fix state 4h. Likewise, the independent properties [T.sub.1h] and [h.sub.3] fix state 4L.

The low temperature evaporator cooling load, [Q.sub.L], and the enthalpies at state 1L and4L are used to calculate the ammonia mass flow rate through the low temperature compressor, [m.sub.L].

[m.sub.L] = [Q.sub.L]/([h.sub.1L] - [h.sub.4L]) (3)

Likewise, the high temperature evaporator cooling load, [Q.sub.h], and the enthalpies at state 1h and 4h are used to calculate the mass flow rate through the high temperature compressor, [m.sub.h].

[m.sub.h] = [Q.sub.h]/([h.sub.1h] - [h.sub.4h]) (4)

The low temperature compressor power, W[c.sub.L], is calculated using the enthalpies at states 1L and 2L and the mass flow rates through the low temperature compressor as:

W[c.sub.L] = [m.sub.L] x ([h.sub.2L] - [h.sub.1L]) (5)

Similarly, high temperature compressor power, W[c.sub.h], is calculated using the enthalpies at states 1h and 2h and the mass flow rates through the high temperature compressor as:

W[c.sub.h] = [m.sub.h] x ([h.sub.2h] - [h.sub.1h]) (6)

The total power draw for both compressors, W[c.sub.tot], is calculated using the motor efficiency, [[eta].sub.m], as:

W[c.sub.tot] = (W[c.sub.L] + W[c.sub.h])/[[eta].sub.m] (7)

Evaporative Condenser Model

The effectiveness, eff, of an evaporative condenser can be modeled as (Manske et al., 2001):

eff = Actual Condenser Capacity/Maximum Condeser Capacity

eff = [[??].sub.air] x ([h.sub.air,out] - [h.sub.air,in])/[[??].sub.air] x ([h.sub.air,out][[parallel].sub.Trefgerant.Sat] - [h.sub.air,in]) (8)

where:

[[??].sub.air] = air mass flow rate through evaporative condenser

[h.sub.air,out] = enthalpy of air leaving evaporative condenser

[h.sub.air,in] = enthalpy of air entering evaporative condenser, enthalpy of air at out door air conditions

[h.sub.air,out][[parallel].sub.Trefgerant,Sat] = enthalpy of saturated air at ammonia saturation temperature

As indicated by Equation 8, the air mass flow rate and the actual condenser capacity for the given outdoor air conditions are required to determine the effectiveness of the evaporative condenser. Typically, manufacturers report volume air flow rate, nominal capacity, and the heat rejection factor (HRF) (Manske et al., 2001). The volume air flow rate is used to calculate the air mass flow rate using the density of air at standard conditions. The HRF, which is a function of both the outside air wet bulb and the saturated condensing temperature (SCT or [T.sub.3]), is used to determine the condenser capacity for a given outside wet-bulb temperature and SCT as (Manske et al., 2001):

Condenser [Capacity.sub.Actual] = Nominal Capacity/HRF(Twb, SCT) (9)

Equations 8 and 9 were applied to the specifications listed for a typical forced-counter flow, evaporative condenser (Evapco, 2010) to determine the relationship between SCT and effectiveness for a wet-bulb range of 50[degrees]F to 86[degrees]F (10[degrees]C to 30[degrees]C). Effectiveness was found to be linearly related to SCT with an [R.sup.2] of 0.94 as:

eff = 1.0231 - 5.2E-3 x SCT (10)

The required condenser capacity of the refrigeration system must also be determined to incorporate the evaporative condenser model through use of Equations 8 and 10. The condenser capacity can be calculated as:

Cond Capacity Required = [m.sub.L] x ([h.sub.2L] - [h.sub.3]) + [m.sub.h] x ([h.sub.2h] - [h.sub.3]) (11)

Equations 8, 10 and 11 are incorporated into the refrigeration model, all mass flow rates and state properties can be determined by solving the previous set of equations.

Most refrigeration control sequences maintain a minimum head pressure setpoint. As long as the head pressure is above this setpoint, the evaporative condenser fans and pumps run at full capacity. When the head pressure drops below this setpoint, the evaporative condenser fans are modulated to maintain the minimum head pressure. Fraction load ([FL.sub.fans]) on the fans can be calculated as:

[F.sub.Lfans] = Cond [Cap.sub.Required]/Cond [Cap.sub.Actual] (12)

If the condensers do not have enough condenser capacity to maintain the setpoint, the fans run at full speed and an iterative method is used to calculate the head pressure. Hence an "if" statement is required to make the logic work.

Based on the fan affinity laws and accounting for decreased motor efficiency at reduced speeds, the actual power, [P.sub.act], can be calculated from the power at full fan flow, [P.sub.full], and fan motor efficiency, [[eta].sub.m], as:

[P.sub.act] = ([P.sub.full] x [([FL.sub.fans]).sup.2.5])/[[eta].sub.m] (13)

Heat Reclaim Model

In addition to the evaporative condenser model, a heat exchanger model can also be added to the base refrigeration models to reclaim heat from the compressor discharge. For the single-stage refrigeration system, an energy balance on the discharge header is performed to calculate the mixed ammonia enthalpy, [[eta].sub.m], as a result of mixing states 2L and 2h. For entering mass flow rates and enthalpies, the mixed enthalpy, [[eta].sub.m], is calculated as:

[h.sub.2m] = ([m.sub.L] x [h.sub.2L] + [m.sub.h] x [h.sub.2L])/([m.sub.L] + [m.sub.h]) (14)

However, not all the heat from point 2m is available for reclaim since some heat is removed from the refrigerant through lubricating oil or cylinder head cooling in reciprocating compressors. Thus, refrigerant exit temperatures are on the order of 175[degrees]F to 195[degrees]F (80[degrees]C to 90[degrees]C) for reciprocating compressors and 120[degrees]F to 130[degrees]F (48[degrees]C to 54[degrees]C) for screw compressors (Stoecker, 1998). This heat loss can be modeled by assuming that the discharge from point 2m passes through a heat exchanger and loses some heat to the oil. [Q.sub.oil], is calculated as:

[Q.sub.oil] = [eff.sub.hx_oil] x ([m.sub.oil] x C[p.sub.oil]) x ([T.sub.2m] - [T.sub.oil,in]) (15)

Where, [eff.sub.hx_oil], is the heat exchanger effectiveness and T2m is the temperature of superheated refrigerant, [m.sub.oil] is mass flow rate of oil, [Cp.sub.oil] is specific heat of oil at constant pressure and [T.sub.oil,in] is the inlet temperature of oil. Typical oil temperature for old screw compressors is about 130[degrees]F (55[degrees]C). In this approximate approach heat exchanger effectiveness, the mass flow rate and inlet temperature of the oil are set to achieve the exit refrigerant temperatures noted above. The new discharge temperature can be calculated as:

[T.sub.2m_new] = [T.sub.2m] - [Q.sub.oil]/([m.sub.oil] x [Cp.sub.oil]) (16)

Assuming the heat exchanger is used for heating water and is sized to only extract heat from the superheated region, the heat extracted by the heat exchanger, [Q.sub.hx], is calculated as:

[Q.sub.hx] = [eff.sub.hx] x ([m.sub.water] x [Cp.sub.water]) x ([T.sub.2m_new] - [T.sub.w,in]) (17)

Where, [eff.sub.hx], is the heat exchanger effectiveness and T2m new is the new temperature of superheated mix, [m.sub.w] is mass flow rate of water, [Cp.sub.w] is specific heat of water at constant pressure and [T.sub.w,in] is the inlet temperature of water.

The pressure drop from the heat exchanger is assumed to be zero. The heat exchanger will also decrease the amount of heat the evaporative condenser must reject. Thus, for a system with a sensible-heat heat exchanger, Equation 11 would be modified to:

Cond [Capacity.sub.Required] = [m.sub.L] x ([h.sub.2L] - [h.sub.3]) + [m.sub.h] x ([h.sub.2h] - [h.sub.3]) - [Q.sub.hx] (18)

BASE CASE SYSTEM

Using the method developed above, a single-stage ammonia refrigeration system is modeled as a base case to evaluate energy saving opportunities. The base case system has two refrigeration circuits: a) process load of 200 tons (703 kW) that remains constant throughout the year and b) load from the cold storage rooms that varies linearly from 600 tons (2,110 kW) in summer to 400 tons (1,407 kW) during the winter. Both circuits operate at a suction pressure of 20 psig (240 kPa) corresponding to an evaporating temperature of 6[degrees]F (-15[degrees]C). The minimum system pressure setpoint is currently set at 140 psig (1,066 kPa), but the system could run as low as 110 psig (860 kPa).

The base case system uses reciprocating compressors equipped with cylinder unloading capabilities, with the assumption that the fraction of full-load power matches the fraction of full-load capacity. The isentropic efficiencies of the compressors are assumed to be constant at 75%. The evaporative condenser capacity is 26,400 MBH (7,737 kW). The evaporative condenser capacity was sized to meet a cooling load of 800 tons (2,814 kW) at a peak wet-bulb temperature of 80[degrees]F (26.7[degrees]C) with a maximum head pressure of 175 psig (1,308 kPa), but has developed scaling of about 1/32 inch (0.8mm) over time. The evaporative condenser has a 90 hp (68 kW) fan with a 20 hp (15 kW) water pump. The evaporative condenser fans are controlled with VFDs that slow the fans when the minimum head pressure setpoint is met. The mass flow rate of air through the condenser is 352,000 cfm (166 [m.sup.3]/s). All motors are assumed to be 90% efficient. The pumps are turned off when the wet-bulb temperature is at or below 32[degrees]F (0[degrees]C). The model includes compressor power, evaporative condenser fan and pump power but does not include evaporator fan power.

Figure 2 shows the variation of system power draw and head pressure with ambient wet-bulb temperature for the base case system. The slope of system power changes at 55[degrees]F (13[degrees]C) wet-bulb temperature. This temperature change point is caused by the minimum head pressure setpoint. The head pressure increases above 55[degrees]F (13[degrees]C) wet-bulb temperature because a higher condensing temperature and pressure is required to reject system heat to the environment. The head pressure remains constant below 55[degrees]F (13[degrees]C) wet-bulb temperature because the minimum head pressure setpoint is being maintained. This means that below the 55[degrees]F (13[degrees]C) wet-bulb temperature, compressor power per ton of refrigeration load remains constant and the reduction in system power is due to the reduction in the refrigeration load. Close inspection of the graph shows a small reduction in total power below 32[degrees]F (0[degrees]C) wet-bulb temperature because evaporative condenser pumps are turned off at low wet-bulb temperatures. The head pressure is higher than the design conditions because the evaporative condenser has developed scaling.

ENERGY EFFICIENCY OPPORTUNITIES

Energy consumption of an ammonia refrigeration system primarily depends on refrigeration load, refrigeration temperature, evaporative condenser control, evaporative condenser effectiveness along with some other secondary parameters. Thus principal methods of reducing energy use and improving energy efficiency are:

1. Reduce refrigeration load

2. Increase suction pressure

3. Employ dual suction pressures

4. Reduce minimum head pressure setpoint

5. Improve evaporative condenser effectiveness

6. Reclaim heat

The following sections describe the energy response of the base case system to these measures, and describe some methods of how these measures can be applied.

Reduce Refrigeration Load

Reducing the refrigeration load reduces the power consumption of the system. Reduced refrigeration load can be achieved by: a) insulating cold spaces and refrigerant piping, b) reducing infiltration by sealing doors and controlling door operations, c) upgrading to more efficient lighting in refrigerated spaces, d) reducing hot gas defrost to a required minimum, and e) adding VFDs on evaporator fan motors.

The actual reduction in compressor power with load also depends on the part-load performance of the compressor(s). In this analysis, we assume that the base case system uses reciprocating compressors equipped with cylinder unloading capabilities and that the isentropic efficiencies of the compressors are constant at 75%. With these types of compressors and controls, the fraction of full-load power nearly matches the fraction of full-load capacity, giving excellent part-load performance. Screw compressors with a combination of slide valve and variable-speed control may also have excellent part-load performance. In other types of compressors and controls, such as screw compressors with modulation control, the fraction of full-load power does not fall as quickly as fraction of full-load capacity, and part-load performance is not as good.

Figure 3 shows power draw of the base-case system and power draw of the system assuming the refrigeration load is reduced by 50 tons (176 kW) on the second circuit by moving to more energy efficient lighting and reducing infiltration loads. Note that power draw from reducing the refrigeration load is reduced over the entire range of ambient wet-bulb temperatures.

Increase Suction Pressure

Suction pressure should be set to maintain the required temperature of the refrigerated product or space. A lower suction pressure increases the compression ratio that the compressor must develop resulting in more work per refrigeration load. To determine the suction pressure during the design stage, determine the maximum safe temperature for the product or space, then determine the required temperature difference between the temperature of the product or space and the evaporating temperature of the refrigerant. In existing systems, one indication of excessively low suction pressure is frequent cycling of the evaporator fans. In this case, suction pressure can simply be increased. Suction pressure may also be increased by: a) replacing evaporators with new evaporators with higher heat exchanger effectiveness, b) adding more evaporators to increase the available surface area, c) replacing undersized suction piping, and d) improved maintenance and cleaning of evaporator surfaces. Raising the suction pressure is likely to increase the fan energy use on the evaporator side, but net energy use is nearly always reduced since compressor energy use decreases more than fan energy use increases.

Figure 4 shows power draw of the base-case system and power draw of the system assuming the suction pressure is raised on both circuits by 5 psi (34.5 kPa) to 25 psig (274 kPa), with a corresponding evaporating temperature of 11[degrees]F (-12[degrees]C). Note that power draw from increasing suction pressure is reduced over the entire range of ambient wet-bulb temperatures.

Employ Dual Suction

In some facilities, different products or spaces must be maintained at different temperatures. In these cases, power use can be reduced by maintaining different suction pressures for the different products or spaces rather than maintaining a single suction pressure for the lowest temperature requirement in the system. Doing so requires separate sets of compressors to maintain the dual suction pressures. The resulting "dual-suction" system has a high-suction loop, and a low-suction loop with a common head pressure as illustrated in Figures 1a and 1b. In facilities with a single suction system the suction pressure is dictated by the lowest temperature requirement in the system, which is only a fraction of the total load. Most facilities already have multiple compressors. In some cases dual suction can be employed with only a few modifications to the piping and control system.

Figure 5 shows power draw of the base-case system and power draw of the system assuming the suction pressure could be raised on the second circuit to 40 psig (377 kPa), with a corresponding evaporating temperature of 26[degrees]F (-3[degrees]C). Note that power draw from increasing suction pressure is reduced over the entire range of ambient wet-bulb temperatures.

Reduce Minimum Head Pressure Setpoint

Most refrigeration systems maintain a minimum head pressure setpoint for compressor oil cooling and/or hot gas defrost. In most systems, the actual head pressure only approaches the minimum head pressure setpoint at moderate and low ambient wet-bulb temperatures. When the actual head pressure reaches the minimum head pressure setpoint, evaporative condenser fans slow down to maintain this setpoint when controlled with VFDs.

Reducing the minimum head pressure setpoint reduces power consumption at moderate and low ambient wet-bulb temperatures. In some cases, the minimum head pressure setpoint is well above the minimum head pressure actually required by the system and can simply be reduced. In other cases, where compressor oil cooling is achieved by using the refrigerant (thermosyphon or direct injection), systems using direct refrigerant injection oil cooling require higher head pressures than systems using thermosyphon oil cooling because the pressure differential drives the oil cooling process; in these systems, switching to a more efficient oil cooling process like thermosyphon oil cooling or replacing the thermostatic expansion (TXV) valves with electronic expansion valves can make it possible to reduce the minimum pressure setpoint. In other cases, minimum head pressure is dictated by the requirements for delivery of high pressure liquid to direct expansion loads and hot gas defrost. In the latter case, replacing undersized hot gas piping would help; otherwise computer controls may be able to raise the head pressure during defrost.

Figure 6 shows power draw of the base-case system and power draw of the system assuming the minimum head pressure setpoint is reduced from 140 psig (1,066 kPa) to 110 psig (860 kPa). Note that power draw from reducing minimum head pressure setpoint is reduced only during moderate and low ambient wet-bulb temperatures.

Improve Evaporative Condenser Effectiveness

Many plants have underperforming evaporative condensers due to lack of maintenance. For example, spray nozzles may get clogged or dirt and debris may buildup on the condenser acting as insulation. Scale may also build up on the condenser piping; 1/32 inch (0.8 mm) of scale results in 30% capacity loss (Cascade Energy Engineering Inc., 2010). Underperforming evaporative condensers drive the saturated condensing temperature higher in order to reject the heat to the environment. This in turn, raises head pressure and increases compressor power draw. Thus proper evaporative condenser maintenance and condenser water treatment is critical for maximizing energy efficiency.

Figure 7 shows power draw of the base-case system and power draw of the system assuming scale buildup is removed and the evaporative condenser capacity increases by 30% to its original design capacity. Note that power draw from reducing minimum head pressure setpoint is reduced only during moderate and high ambient wet-bulb temperatures.

Reclaim Heat From Compressor Discharge

The temperature of ammonia at the discharge of the compressor usually exceeds 120[degrees]F (49[degrees]C). For facilities with heating demands, reclaiming heat from ammonia at compressor discharge for preheating water should be considered. In addition, reclaiming heat from the compressor discharge reduces the load on the evaporative condenser. Reclaiming heat requires the addition of a heat exchanger and storage tank to store the hot water when hot water demand and refrigeration load are out of phase. The temperature of water in the hot water storage tank is higher than the temperature of the incoming municipal water, which reduces the quantity of heat reclaimed.

Figure 8 shows that the power draw before and after installing a 60% effective heat exchanger with 55[degrees]F (12.8[degrees]C) water from the hot water storage tank. The reduction in system power draw is minimal; however, about 30% of system power is reclaimed as useful heat.

Annual Energy Savings

Figures 3-8 show power requirements over a range of ambient wet-bulb temperatures for base-case systems and systems with energy efficiency opportunities. Annual energy use can be calculated by summing the product of the power draw at each ambient wet-bulb temperature and the number of hours per year at that wet-bulb temperature. Annual energy savings from these energy efficiency opportunities is the difference between the base-case energy use and energy use with the energy efficiency opportunity. If multiple energy efficiency opportunities are implemented simultaneously, the resulting energy savings are typically less than the sum of individual energy savings due to synergistic effects.

To demonstrate this approach, assume the system operates continuously in Dayton, Ohio. Figure 9 shows hours per year at respective wet-bulb temperatures in Dayton, Ohio, obtained using the WeaTran weather data processing software (Kissock, 2011) with TMY3 data (National Renewable Energy Laboratory, 2005).

Table 1 shows the annual system energy use of the base-case system and the system with each energy efficiency opportunity. Note that the sum of the individual savings opportunities is less than the savings if all measures were implemented simultaneously. The savings numbers shown below are specific to the system, climate, and assumptions presented here and would be different for other systems, climates, and assumptions. And certainly, a life cycle cost analysis would be necessary to evaluate the cost-effectiveness of each measure.

SUMMARY AND DISCUSSION

This paper developed a simplified methodology for simulating single-stage ammonia refrigeration systems. It used the methodology to simulate energy use of an example base-case system. It discussed six principles of energy efficient ammonia systems and demonstrated how system power is reduced when each principle is applied. It showed how to use these results to calculate annual energy savings. The magnitude of the savings from each measure depends on many variables; however the results clearly demonstrate that each principle effectively reduces energy use and should be considered when designing new systems or seeking to improve existing systems. Further, a full life cycle cost analysis would be necessary to evaluate the cost-effectiveness of each measure.

General observations about the six principles are:

1. Reducing the refrigerant load by 1% reduces power consumption by 1% in systems in which the fraction of full-load compressor power nearly matches fraction of full-load capacity. In compressors with worse part-load performance, the percentage reduction in power is less than the percentage reduction in refrigeration load.

2. Increasing the suction pressure by 1 psi (6.9 kPa) reduces the power consumption by about 2%

3. Employing dual suction reduces power consumption by the same amount as increasing suction pressure for the fraction of load on which suction pressure is increased.

4. Reducing minimum head pressure setpoint results in power savings at low ambient wet-bulb temperatures.

5. Improving condenser effectiveness results in power savings at elevated ambient wet-bulb temperatures.

6. Installing a heat exchanger to preheat water can reclaim a significant amount of system power as useful heat.

If multiple energy efficiency opportunities are implemented simultaneously, the resulting savings are typically less than the sum of individual savings due to synergistic effects within the system.

The methodology presented here represents an effective method to understand how energy saving principles affect system power in ammonia refrigeration systems. The accuracy of the model could be improved by 1) incorporating compressor part load power characteristics, 2) more detailed modeling of evaporators including evaporator fan control, 3) sequencing and control of multiple condensers. In addition, the scope of the methodology could be broadened to consider other energy saving opportunities including wet-bulb approach control of evaporative condensers, effect of auto-purgers, and effect of subcooling.

ACKNOWLEDGEMENTS

The authors are grateful for support for this work from the U.S. Department of Energy through the Industrial Assessment Center program. The paper was improved from suggestions made by reviewers and others in the ammonia refrigeration industry.

REFERENCES

ASHRAE, 2009, 2009 ASHRAE Handbook--Fundamentals, Chapter 1--Psychrometrics. Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

Cascade Energy Engineering, Inc., 2010, Industrial Refrigeration Best Practices Guide. Chapter 3. Third Edition.

EIA 2006, Manufacturing Energy Consumption Survey (MECS) 2006. http://www.eia.gov/emeu/mecs/.

Evapco, 2010, PMC-E Evaporative Condensers Product Brochure, Bulletin 106C. http://www.evapco.com/products/ pmce_evaporative_condenser.

Kissock, K. 2011, "WeaTran," Version 2011-03-17, http:// academic.udayton.edu/kissock/http/RESEARCH/EnergySoftware.htm.

Manske, K.A., Reindl, D.T., Klein, S.A. 2001. "Evaporative Condenser Control in Industrial Refrigeration Systems." International Journal of Refrigeration. Vol.24, No 7, pg 671-691.

Mathworks Inc., 2012, MATLAB[R] Version: 2012a http:// www.mathworks.com/products/matlab/.

National Renewable Energy Laboratory, 2005, "User's Manual for TMY3s", http://rredc.nrel.gov/solar/old_data/ nsrdb/1991-2005/tmy3/.

National Institute of Standards and Technology, 2010, Reference Fluid Thermodynamic and Transport Properties Database (REFPROP): Version 8.0 http://www.nist.gov/ srd/nist23.cfm.

Stoecker, Wilbert, 1998, Industrial Refrigeration Handbook. McGraw-Hill.

Abdul Qayyum Mohammed

Student Member ASHRAE

Franc Sever, PE

Thomas Wenning

Associate Member ASHRAE

Kelly Kissock PhD, PE

Member ASHRAE

Abdul Qayyum Mohammed is pursuing an MS in Mechanical Engineering at the department of Mechanical and Aerospace Engineering/ Renewable and Clean Energy, University of Dayton, Dayton, OH. Thomas Wenning works as a research engineer at Oak Ridge National Laboratory, Oak Ridge, TN. Franc Sever works as an energy engineer at Go Sustainable Energy, Columbus, OH. Kelly Kissock is a professor and chair of the department of Mechanical and Aerospace Engineering/Renewable and Clean Energy, University of Dayton, Dayton, OH.

Table 1. Energy Use and Savings for the Base-Case System in Dayton, Ohio Energy Efficiency System Electricity Heat Opportunity Energy Use Savings Reclaimed kWh/yr kWh/yr kWh/yr % Base Case System 6,917,555 -- -- -- Reduce Refrigeration 6,400,817 516,738 7.5% -- Load Increase Suction 6,257,762 659,793 9.5% -- Pressure Employ Dual Suction 5,357,237 1,560,318 22.6% -- Pressures Reduce Minimum Head 6,420,410 497,145 7.2% -- Pressure Setpoint Improve Evaporative 6,637,636 279,919 4.0% -- Condenser Effectiveness Reclaim Heat 6,848,438 69,117 1.0% 1,916,442 Sum of Savings for -- 3,583,030 51.8% All Opportunities All Opportunities 4,314,516 2,603,039 37.6% 1,155,702 Together