High-Temperature Heat Pump Model

Documentation and Case Studies

Jordan Cox, Scott Belding, Gustavo Campos, and
Travis Lowder
National Renewable Energy Laboratory

NREL is a national laboratory of the U.S. Department of Energy Technical Report

Office of Energy Efficiency & Renewable Energy NREL/TP-7A40-84560
Operated by the Alliance for Sustainable Energy, LLC August 2023
This report is available at no cost from the National Renewable Energy
Laboratory (NREL) at [Link]/publications.

Contract No. DE-AC36-08GO28308

 High-Temperature Heat Pump Model
Documentation and Case Studies
Jordan Cox, Scott Belding, Gustavo Campos, and
Travis Lowder
National Renewable Energy Laboratory

Suggested Citation
Cox, Jordan, Scott Belding, Gustavo Campos, and Travis Lowder. 2023. High-
Temperature Heat Pump Model Documentation and Case Studies. Golden, CO: National
Renewable Energy Laboratory. NREL/TP-7A40-84560.
[Link]

NREL is a national laboratory of the U.S. Department of Energy Technical Report

Office of Energy Efficiency & Renewable Energy NREL/TP-7A40-84560
Operated by the Alliance for Sustainable Energy, LLC August 2023
This report is available at no cost from the National Renewable Energy National Renewable Energy Laboratory
Laboratory (NREL) at [Link]/publications. 15013 Denver West Parkway
Golden, CO 80401
Contract No. DE-AC36-08GO28308 303-275-3000 • [Link]
 NOTICE

This work was authored by the National Renewable Energy Laboratory, operated by Alliance for Sustainable
Energy, LLC, for the U.S. Department of Energy (DOE) under Contract No. DE-AC36-08GO28308. Funding
provided by the U.S. Department of Energy Office of Energy Efficiency and Renewable Energy Strategic Analysis
Team. The views expressed herein do not necessarily represent the views of the DOE or the U.S. Government.

This report is available at no cost from the National Renewable

Energy Laboratory (NREL) at [Link]/publications.

U.S. Department of Energy (DOE) reports produced after 1991

and a growing number of pre-1991 documents are available
free via [Link].

Cover Photos by Dennis Schroeder: (clockwise, left to right) NREL 51934, NREL 45897, NREL 42160, NREL 45891, NREL 48097,
NREL 46526.

NREL prints on paper that contains recycled content.

Acknowledgments
The National Renewable Energy Laboratory team acknowledges the essential support from the
following individuals and organizations, without whom this work would not have been possible:
Sam Baldwin, U.S. Department of Energy (DOE) Office of Energy Efficiency and Renewable
Energy Strategic Analysis Team; Joe Cresko, DOE Advanced Manufacturing Office; Joe
LaFleur, Commonwealth Dairy; Kegan Grayson and Mark Theisen, Coronado Brewing
Company; Chuck Clerici, Pat Haller, and Melissa Stewart, Efficiency Vermont; and Cordin
Arpagaus and Frédéric Bless, Switzerland Institute for Energy Systems.

iii
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
List of Acronyms
AEC annual energy cost
BAU business as usual
BTU British thermal unit
CO2 carbon dioxide
COP coefficient of performance
FO&M fixed operation and maintenance
HTHP high-temperature heat pump
IRR internal rate of return
LCC life-cycle cost
LCOH levelized cost of heat
NPV net present value
NREL National Renewable Energy Laboratory
PBP payback period
REopt Renewable Energy Integration and Optimization
TEA technical and economic assessment
TOU time of use
UR utility rate
VO&M variable operation and maintenance

iv
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
Executive Summary
High-temperature heat pumps (HTHPs) are electrically powered systems that supply heat above
90°C. HTHPs have the potential to serve two valuable functions in United States (U.S.) industry.
First, by valorizing and elevating waste heat streams, HTHPs can improve industrial energy
efficiency. Second, by electrifying process heat generation, HTHPs can replace combustion
systems with less carbon-intensive sources of energy, such as renewable resources, and can
reduce industrial greenhouse gas emissions. In certain applications, the combination of increased
efficiency, cost savings, and emissions reductions makes HTHPs a promising component of
strategies for industrial electrification and clean energy transitions.

Researchers at the U.S. Department of Energy’s National Renewable Energy Laboratory (NREL)
developed a Python-based HTHP model. This model has both a physics-based performance
estimation component and an economic evaluation component that is designed to demonstrate
the potential economic competitiveness of HTHPs, which was found to be driven by electricity
versus gas prices.

This document has several sections. First, an overview of HTHPs is provided, including a
literature review of relevant state-of-the-art information. Second, a description of the HTHP
model is provided, along with detailed instructions on how to use it. Third, case studies to
showcase the utilization of the HTHP model are provided, along with the corresponding
economic results.

Four major elements comprise the HTHP model’s physics-based component: a coefficient of
performance (COP) calculator for heat pumps given boundary operating conditions, a heat pump
energy balance module, a gas boiler energy balance module, and an electro-resistive energy
balance module. These modules can be combined in a system of heat technologies. An example
is a hybrid heat pump and electro-resistive system. The economics module of the HTHP model
can be used to compare the net present value, the internal rate of return, and the payback period
of any two system configurations. This report provides the equations used for each module and
presents a guide on how to construct a system using the Python programming language for
economic comparison.

To demonstrate the capabilities of the HTHP model, two case studies are presented: beer
brewing and yogurt making. Previous work identified the food and beverage production
industries as the first mover industries for HTHPs (Jakobs and Stadlander 2020). These case
studies were selected because they provide high-temperature heat (>90°C) above what is
provided by traditional heat pumps. Both case studies were evaluated for economic
competitiveness compared to traditional gas heating systems. Our analysis found that the beer-
brewing case is not economically competitive for the given location (California), but the yogurt-
making case is economically competitive for the given location (Vermont).

HTHPs represent a unique opportunity for U.S. industry to decarbonize and recover wasted
energy. This report demonstrates that there are economically competitive opportunities for
HTHPs in the United States. In this work, the case study in Vermont was found to be
economically competitive, though economic competitiveness will depend on the industry and on
electricity versus fossil fuel prices in the local region. In addition to economic competitiveness,

v
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
our partners identified several barriers to HTHP adoption, which are provided in the final section
on discussion and conclusions.

vi
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
Table of Contents
1 Introduction ........................................................................................................................................... 1
1.1 Overview of High-Temperature Heat Pumps ................................................................................ 2
1.2 Types of Heat Pumps .................................................................................................................... 3
1.3 Terms and Key Metrics ................................................................................................................. 5
2 High-Temperature Heat Pump Model Documentation ...................................................................... 7
2.1 Model Physics ............................................................................................................................... 7
2.2 Nomenclature ................................................................................................................................ 7
2.3 Input Libraries ............................................................................................................................... 9
2.4 High-Temperature Heat Pump ...................................................................................................... 9
2.4.1 Single-Stage Coefficient-of-Performance Calculation ..................................................... 9
2.4.2 Single-Stage with Internal Heat Exchanger ................................................................... 11
2.4.3 Single-Stage with Supercritical Fluid............................................................................. 12
2.4.4 Two-Stage Configurations ............................................................................................. 12
2.4.5 Mass and Energy Balance .............................................................................................. 13
2.4.6 Heat Pump Lifetime Costs and Levelized Cost of Heat ................................................. 14
2.4.7 Gas Boiler....................................................................................................................... 15
2.4.8 Electro-Resistive Heater................................................................................................. 16
2.5 Model Economics........................................................................................................................ 16
3 Python Example Inputs ...................................................................................................................... 17
3.1 Default Input Values ................................................................................................................... 17
3.2 Running a Case............................................................................................................................ 19
4 Case Studies of Technical and Economic Potential of Heat Pumps in the United States .......... 20
4.1 Case Study 1: Beer Brewing ....................................................................................................... 20
4.1.1 System Description ........................................................................................................ 20
4.1.2 Model ............................................................................................................................. 22
4.1.3 Inputs .............................................................................................................................. 24
4.1.4 Results ............................................................................................................................ 27
4.2 Case Study 2: Pasteurization and Yogurt Making....................................................................... 30
4.2.1 System Description and Inputs ....................................................................................... 30
4.2.2 Results ............................................................................................................................ 32
5 Discussion and Conclusions ............................................................................................................ 33

vii
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
List of Figures
Figure 1. U.S. energy use (quadrillion BTU) and CO2 emissions by sector in 2019 .................................... 1
Figure 2. Cumulative distribution of U.S. industrial heat demand by temperature ...................................... 2
Figure 3. Schematic of a mechanical heat pump........................................................................................... 3
Figure 4. Heat pump thermodynamic cycle .................................................................................................. 6
Figure 5. Temperature-entropy diagram ....................................................................................................... 8
Figure 6. Single-stage cycle with internal heat exchanger .......................................................................... 11
Figure 7. Supercritical cycle thermodynamic diagram and effect of pressure ratio on COP ...................... 12
Figure 8. Two-stage configurations: cascade, splitter economizer, flash economizer, and flash intercooler
................................................................................................................................................ 13
Figure 9. Beer brewing process................................................................................................................... 21
Figure 10. Heat consumption by source (left) and electricity and natural gas consumption by source
(right) ..................................................................................................................................... 21
Figure 11. Two-stage heat pump cycle using R245fa and splitter/mixer between stages........................... 23
Figure 12. HTHP model COP compared to temperature lift (sink temperature fixed at 120°) .................. 24
Figure 13. Brewery monthly energy consumption for 1 year from an SDG&E utility bill. Peak demand
data are available only for the month of June 2022 and for the yearly maximum. ................ 25
Figure 14. Wort boiling thermal load profile (operation schedule of the heat pump) ................................ 26
Figure 15. TOU electricity prices for AL-TOU2 from SDG&E ................................................................. 26
Figure 16. Monthly natural gas prices for 1 year ........................................................................................ 27
Figure 17. Operating costs of the variable and fixed components of the electric heat pump and natural gas
boilers ..................................................................................................................................... 28
Figure 18. Impact of capital and operating costs on the LCC of the electric heat pump and natural gas
boilers ..................................................................................................................................... 28
Figure 19. Effect of COP and natural gas price on equipment LCC........................................................... 29
Figure 20. Battery energy storage optimal dispatch profile ........................................................................ 30
Figure 21. Commonwealth Dairy daily thermal energy demand ................................................................ 31

List of Tables
Table 1. Fundamental Types of Heat Pumps ................................................................................................ 4
Table 2. Example Process Data From Input Libraries .................................................................................. 9
Table 3. Default Input Values ..................................................................................................................... 18
Table 4. Specifications of Kobelco Steam Generating Heat Pumps (SGH120 and SGH165).................... 22
Table 5. Experimental Results for SGH120................................................................................................ 24
Table 6. Input Summary for Beer Brewing Case Study.............................................................................. 27
Table 7. Battery Energy Storage Design Results from NREL’s REopt ...................................................... 29
Table 8. Commonwealth Dairy Input Values ............................................................................................. 31
Table 9. Commonwealth Dairy Key Indicator Results ............................................................................... 32

viii
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
1 Introduction
The industrial sector is a major energy consumer and a source of carbon dioxide (CO2) emissions
in the United States, accounting for approximately 35% of total energy end-use consumption and
28% of CO2 emissions in 2019 (Figure 1). The industrial sector is considered difficult to
decarbonize because of its high reliance on fossil fuels, processes that cannot be directly replaced
by renewable resources, and a lack of robust research on electrification methods (Madeddu
2020). Process heat (thermal energy used in the manufacturing of products) comprises the
majority (51%) of industrial energy demand in the United States (Ruth 2019). Importantly, more
than two-thirds of this heat demand are for processes at or below 300°C (Figure 2). Further, of
the more than 7 quadrillion British thermal units (BTUs) demanded annually for process heating
in the United States, 36% is lost as unrecovered waste heat, leading to 2.5 quadrillion BTUs of
heat with the potential to be harnessed through increased efficiency (DOE 2015).

Figure 1. U.S. energy use (quadrillion BTU) and CO2 emissions by sector in 2019
Source: EIA 2020. Figure by NREL

High-temperature heat pumps (HTHPs) available today can provide heat at temperatures
between 90°C and 160°C. (Some models can go higher but have lower performance and thus
were not the focus of this work (Arpagaus et al. 2018)). Process heat in this range can be useful
to many industrial sectors, including food and beverage manufacturing, corn milling, pulp and
paper production, industrial drying, and the chemical industry. For widespread sectoral adoption,
however, HTHPs must present a compelling economic case, achieved through energy cost
savings. The effectiveness of a heat pump can be expressed through the coefficient of
performance (COP), which describes the HTHP’s thermal energy output in relation to its
demanded electrical energy. Due to thermodynamic limits and refrigerant and compressor
inefficiencies in real-world operation, most current HTHPs operate within an approximate COP
range from 2–6, or between 40%–50% of their theoretical maximum (Carnot) efficiency

1
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
(Arpagaus et al. 2018; Arpagaus and Bertsch 2020; Kosmadakis et al. 2020; Schlosser et al.
2020; Tveit 2017). HTHP performance levels required for economic viability are highly
contextual because of adoption barriers and facility-scale conditions.

Despite these barriers and uncertainties, HTHPs are well matched to many applications, and
active areas of research can improve the understanding, performance, and reliability of HTHPs to
fully realize the full benefits of industrial heat pump adoption.

Circumstances such as high natural gas prices and decreasing electricity costs from renewable
resources present large economic and decarbonization opportunities via both waste heat capture
and the provision of industrial process heat. HTHPs are a promising technology on both fronts.
“High” is a relative term; HTHPs include heat pumps that deliver temperatures above 90°C and
historically have been limited to a maximum of roughly 160°C. Although this range includes the
extreme frontier of what heat pumps as a technology have been able to deliver, these
temperatures are still on the low end of the industrial use spectrum. Because heat pumps elevate
an existing waste/available heat stream, they can improve the usefulness and increase the
utilization of industrial waste heat. Also, heat pumps are electric, which means they possess the
potential (depending on grid generation sources) to decarbonize industrial energy by displacing
fossil-fueled combustion systems, such as natural gas-fired boilers.

Figure 2. Cumulative distribution of U.S. industrial heat demand by temperature

Source: McMillan 2019. Used with permission

1.1 Overview of High-Temperature Heat Pumps

Although heat pumps employ varied technologies, configurations, and working fluids, all heat
pumps perform the same basic functions (DOE 2014):

• Receipt of heat from a source (ambient or waste heat)

• Increase of source heat temperature
• Delivery of useful heat at an elevated temperature.

2
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
Mechanical heat pumps use phase changes to affect this temperature increase and are common in
the space heating sector, among others. Compounds that undergo these phase changes are
referred to as working fluids or refrigerants. Open-cycle heat pumps use the same fluid that
carries and delivers process heat as their working fluid as well (usually water). Closed-cycle heat
pumps have a closed working fluid or refrigerant loop. In both cases, the waste heat source
evaporates the working fluid, converting it from a liquid to a gas. The supplied electricity is
converted into work at the compressor, which increases the pressure on the working fluid and
causes it to condense into a liquid at the condenser. This condensation phase change releases
useful, high-temperature heat to the destination process stream, or heat sink. The working fluid is
then allowed to expand back to the evaporator, where the cycle begins anew (Figure 3).

Figure 3. Schematic of a mechanical heat pump

Figure by NREL

1.2 Types of Heat Pumps

The two main categories of HTHP technologies are vapor compression and absorption (Frate et
al. 2019). A variety of compressor types can be found in HTHPs (see Table 1) and most
commercial models use vapor compression technology, and thus, correspondingly, this document
primarily focuses on such technologies. Absorption heat pumps typically require higher capital
investment per unit of heating capacity, cannot modulate heat output as flexibly as compression
pumps caused by their large active thermal mass, require high-temperature waste heat streams to
deliver heat above 90°C, and lack the capacity to effectively integrate variable energy sources
(Wolf and Blesl 2016; David 2017). Other innovative models center on heat- or chemically

3
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
driven heat pump cycles and have fewer reference points in the market with which to predict
future adoption (Verdeyen 2017). Table 1 describes the primary types of heat pumps (DOE
2014; Laue 2014; Liew 2016; Jianyong 2014).
Table 1. Fundamental Types of Heat Pumps

Heat Pump Type Description Advantages Drawbacks

Matching
refrigerant and
Described in the representative Flexibility in terms of
compressor
schematic in Figure 3, closed- compressor design,
properties with
cycle mechanical heat pumps refrigerant selection,
Mechanical, high-temperature
impart heat followed by pressure and combination
closed cycle conditions can be
to an isolated working fluid, with linked stages to
difficult to
whose condensation allows the allow for larger
technically and
recovery of upgraded heat. temperature lifts
economically
optimize.
Also known as mechanical vapor
recompression, open-cycle A natural fit for
Must use water
mechanical heat pumps use processes that use
as the working
water as a working fluid and can steam and require
Mechanical, fluid; subject to
directly use steam by temperature lifts that
open cycle lower efficiency at
compressing it to a higher fall within water’s
high-temperature
pressure and using the energy capabilities as a
lifts
from this condensation to supply refrigerant
heat to a process.
Higher capital
High-temperature costs, less
Absorption systems combine two
lifts are possible, flexible output
input heat streams, one of low-
high performance modulation,
grade waste heat and another of
can be achieved at requires presence
high-grade heat, to output a heat
Absorption, high-temperature of high-
stream with a useful temperature
closed cycle lifts, and it can temperature
somewhere between those of the
accommodate waste heat
two input streams. This entails a
simultaneous streams, difficult
four-component heat exchanger
heating and cooling to optimize
system.
applications. working pair
selections
Also known as steam jet ejectors,
thermocompressors take in low-
pressure suction steam and Must use water
compress it to a more useful High-pressure as the working
Thermocompression, high-pressure level. The steam lifts low-grade fluid; subject to
open cycle compression agent is the ejector, waste vapor to a lower efficiency at
which uses a tapered nozzle, useful temperature. high-temperature
mixing chamber, and diffuser to lifts
combine, propel, and pressurize
two vapor streams.

4
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
1.3 Terms and Key Metrics
Temperature lift is the difference between the temperature of a heat pump’s source heat stream
(source temperature (𝑇𝑇𝐿𝐿 )) and the temperature of the supplied heat output (sink temperature
(𝑇𝑇𝐻𝐻 )). Temperature lift is a key determinant of ultimate heat pump performance and is an
important metric for evaluating industrial processes and the potential for HTHP integration.

Temperature glide refers to a refrigerant’s evaporation or condensation occurring across a range

of temperatures rather than at a single temperature (holding pressure constant). Using a mixture
of multiple different refrigerants (a zeotropic mixture) can impose these temperature glide
conditions upon a heat pump cycle. When optimized, these glides can decrease losses because of
heat transfer and result in a more efficient HTHP; however, if the temperature glides are
mismatched to the heat pump cycle caused by insufficient design considerations or operational
leakage that alters the refrigerant mixture’s composition, the heat pump’s performance can be
adversely affected (Zühlsdorf et al. 2018).

The coefficient of performance (COP) expresses the ratio of heat supplied by the pump (Qh) to
the input power required by the compressor (W). The higher the COP, the more heat that can be
supplied per unit of electricity, and the more effective the heat pump. A heat pump with a
suitably high COP can provide cost savings compared to competitor heat-provision technologies.

𝑄𝑄ℎ
𝐶𝐶𝐶𝐶𝐶𝐶 =
𝑊𝑊

The COP is limited to a theoretical maximum (Carnot) efficiency based on the temperatures (in
Kelvin) at the heat pump’s condenser (hot side or sink) (TH) and evaporator (cold side or source)
(TC). For heat pumps that involve temperature glides or transcritical cycles, 1 performance is
limited by Lorenz efficiency rather than Carnot efficiency, which accounts for these conditions
using the logarithmic mean temperature in the gas cooler (Tm). Thermodynamics dictate lower
possible COPs for larger temperature lifts (𝐶𝐶𝐶𝐶𝐶𝐶 ∝ 1/∆𝑇𝑇).
𝑇𝑇𝐻𝐻 𝑇𝑇𝑚𝑚
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 𝐶𝐶𝐶𝐶𝐶𝐶𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 =
𝑇𝑇𝐻𝐻 −𝑇𝑇𝐶𝐶 𝑇𝑇𝑚𝑚 −𝑇𝑇𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒

Theoretical Carnot and Lorenz efficiencies, however, are not achieved in real-world heat pumps
as a result of refrigerant performance, exergy loss, and irreversibility/efficiency losses
throughout the system. In practice, HTHPs can achieve 30%–65% of the theoretical maximum
COP, though most industrial HTHPs fall between 40%–50% (Schlosser et al. 2020). Estimating
the real-world efficiency of HTHPs will require additional experimentation, but for economy-

1
Transcritical heat pump cycles do not achieve heat rejection through a phase change in a subcritical region of the
working fluid. Instead, transcritical cycles perform evaporation in a subcritical region and gas cooling/heat rejection
in a supercritical region. CO2 (R744) is the most common refrigerant for transcritical heat pump cycles. Its
thermodynamic properties are such that a conventional subcritical phase change is a low-efficiency heat transfer
process, but CO2 vapor compressed beyond its critical pressure can deliver much higher heat rejection enthalpy
(Rony 2019).

5
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
wide analysis, a Carnot efficiency factor (𝜂𝜂) can be used to estimate real-world performance
from the theoretical efficiency.
𝑇𝑇𝐻𝐻
𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 ≈ 𝜂𝜂 × 𝐶𝐶𝐶𝐶𝑃𝑃𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 = 𝜂𝜂 ×
𝑇𝑇𝐻𝐻 − 𝑇𝑇𝐶𝐶
Because the compressor is the largest source of parasitic energy, a compressor’s efficiency (𝜀𝜀)
can also be used to estimate the actual versus ideal Carnot efficiency. The compressor work is
estimated as the change in the enthalpy of the working fluid across the compression phase of the
vapor cycle on a mass-flow basis (kJ/kg), as shown in Figure 4.

𝑄𝑄̇ℎ (ℎ3 − ℎ2 )
𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 ≈ ≈
𝑊𝑊̇𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 1
𝜀𝜀 (ℎ2 − ℎ1 )

Figure 4. Heat pump thermodynamic cycle

Source: figure by NREL

Heat pump capacity typically refers to the thermal capacity output of an HTHP (kWth). This is
important to differentiate from the electrical capacity demanded by the HTHP. Although these
two values are connected through the COP, each have unique considerations. Thermal capacity is
required for correctly matching an HTHP to its application and process needs, whereas electrical
capacity is a primary driver of variable operational costs associated with HTHP operation.

6
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
2 High-Temperature Heat Pump Model Documentation
It is intended that the heat pump model is open source; it is currently available on GitHub at:
[Link]

2.1 Model Physics

The HTHP technical and economic model provided here was built by the project team to
accomplish the following objectives:

• Establish a library of industrial processes that could use industrial heat pumps.
• Identify potential refrigerants, estimate compressor efficiency, and estimate heat pump
COP for the industrial processes in the industrial process library.
• Estimate the technical and economic competitiveness of HTHPs compared to natural gas
systems by calculating the lifetime costs, levelized cost of heat (LCOH), net present
value (NPV), internal rate of return (IRR), and payback period (PBP) for given industrial
processes. The model can compare heat pumps to new natural gas systems, existing
natural gas systems, and future scenarios involving a price on CO2 emissions.
• Verify and validate the model against published literature and industrial data.
• Use the validated model to estimate the areas that might be of highest interest for more
detailed technical and economic assessment (TEA).
The model can be divided into six main modules meant to support these objectives:

1. A library of inputs that can be used to auto-populate the heat pump model with cold-side
process streams, hot-side process streams, potential refrigerants to be tested, and potential
compressors to be tested.
2. A COP calculator that estimates a theoretical and actual COP as well as suggests
refrigerants and compressor technologies to be used for the heat pump.
3. A mass and energy balance calculator that estimates the mass flow of the hot- and cold-
side process streams.
4. A heat pump economics calculator that uses the process heat requirement, heat pump
performance, and estimated costs to calculate the life-cycle costs (LCCs) of the heat
pump and the LCOH.
5. A natural gas economics calculator that reproduces the estimated LCCs for the same
process driven by a natural gas boiler and estimates natural gas emissions.
6. A cash flow model that uses the operating and fuel savings of the heat pump versus the
natural gas system to calculate the NPV and IRR of a heat pump for the industrial
process.

2.2 Nomenclature
Heat pump literature uses several different terms, including “working fluid,” “heat sink,” and
“heat source.” For the readability of this model, the authors chose to use as simple names as
possible. All units of power are electric unless indicated by the addition of “th,” meaning
“thermal.” For this work, the systems are described below.

The heat exchanger that dumps heat into the industrial process (the condenser) is referred to as
the “hot side” because it will always be the highest temperature. (In the literature, this is

7
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
sometimes referred to as the “heat sink.”) The heat transferred to the hot side is 𝑞𝑞ℎ and is on a
per-mass basis and 𝑞𝑞ℎ is equal to the change in enthalpy per-mass of the refrigerant between
points 2 and 3, indicated in Figure 3, Figure 4, and Figure 5. Enthalpy is denoted as the variable
ℎ.

𝑞𝑞ℎ = (ℎ3 − ℎ2 )

The heat exchanger (evaporator) that pulls heat from the waste or ambient stream is referred to as
the “cold side” because it will always be the coldest temperature (sometimes referred to as the
“heat source”). The heat transfer from the cold side is 𝑞𝑞𝑐𝑐 and is on a per-mass basis and 𝑞𝑞𝑐𝑐 is
equal to the change in enthalpy of the refrigerant between points 4 and 1, indicated in Figure 3
and Figure 5.

𝑞𝑞𝐶𝐶 = (ℎ1 − ℎ4 )

The “process stream” always refers to the fluid outside of the heat exchangers and is labeled
either “hot process stream” or “cold process stream,” correlating to the hot and cold sides,
respectively. The fluid inside the heat pump, sometimes known as the “working fluid,” is
hereafter referred to as the “refrigerant,” unless the heat pump is an open cycle, where the heat
transfer fluid and refrigerant are the same. The term refrigerant is used because of its history of
use in refrigeration cycles, which are commonly heat pumps run in reverse.

The compressor is the primary driver of the heat pump and the major source of energy
consumption. The per-mass work of the compressor (𝑤𝑤𝑐𝑐 ) is expressed as a ratio of the enthalpy
change between points 1 and 2 and the compressor’s estimated isentropic efficiency (𝜀𝜀𝑐𝑐 𝜀𝜀𝑣𝑣 ) 2.

(ℎ2 − ℎ1 )
𝑤𝑤𝑐𝑐 =
𝜀𝜀𝑐𝑐 𝜀𝜀𝑣𝑣

𝜀𝜀𝑐𝑐 = 𝑓𝑓(𝑃𝑃, 𝑑𝑑𝑑𝑑, 𝑇𝑇, 𝑑𝑑𝑑𝑑, 𝑚𝑚̇)

These points correlate to the temperature-entropy diagram shown in Figure 5.

Figure 5. Temperature-entropy diagram

Source: Whitman 2020

2
An ideal compressor has no losses from electrical or mechanical inefficiencies (such as friction). Isentropic
efficiency is the relationship of ideal vs. real power needed in a compressor. Here this is expressed in two variables
multiplied, the compressor efficiency (𝜀𝜀𝑐𝑐 ) and volume ratio efficiency (𝜀𝜀𝑣𝑣 ).

8
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
2.3 Input Libraries
To enable the examination of many industries and heat pump technologies, input libraries were
built that include industrial processes and their operating temperatures and process streams,
refrigerant properties, and compressor properties. A full list of the industrial processes,
temperatures, process streams, refrigerants, and compressors are included with the Jupyter
workbook provided in the GitHub repository. Table 2 shows sample data reported in these
libraries. “High temperature needed” is the maximum temperature needed by the process, and
“hot temperature minimum” is the minimum allowable temperature for the hot working fluid.
(This is the temperature at which the hot working fluid reenters the heat pump for reheating.)

Table 2. Example Process Data From Input Libraries

Hot Waste
High
Heating/ Process Temperature Temperature
Process Name Temperature
Cooling Stream Minimum Available
Needed (°C)
(°C) (°C)
Pasteurization
Heating Water 80 75 50
(low)
Pasteurization
Heating Water 90 85 50
(high)
Concentration
Heating Water 70 60 50
(w/o boiling)
Yogurt Heating Water 95 90 50
Water heating Heating Water 55 50 15
Cleaning Heating Water 85 60 15

2.4 High-Temperature Heat Pump

Each module is governed by equations that are stated in the following sections. This section
describes each formula and provide a brief description of the nomenclature.

2.4.1 Single-Stage Coefficient-of-Performance Calculation

The COP can be calculated or given based on the goal of a TEA. For example, if the goal of a
given TEA is to understand the break-even COP or the COP where the LCOH from a heat pump
is less than that of a gas system, the model can take in an artificial range of COPs (i.e., [1,6]) and
estimate the break-even point. If the model is given only industrial processes or temperature
ranges, it will attempt to calculate the COP in one of two ways, as described below. For the
following methods, note that there are no perfect heat exchangers. This means that the refrigerant
temperature must be higher or lower than the hot-side or cold-side working fluids (respectively)
to facilitate heat transfer; therefore, the authors chose a temperature difference of 5°C. This
difference is often referred to as the “approach temperature” of the heat exchanger.

𝑇𝑇𝐻𝐻 = 𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 ℎ𝑖𝑖𝑖𝑖ℎ + 5°𝐶𝐶 and 𝑇𝑇𝐶𝐶 = 𝑇𝑇𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 − 5°𝐶𝐶

9
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
Method 1: Carnot Factor

The ideal Carnot COP is based on the temperature lift and is given as:

𝑇𝑇𝐻𝐻
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 =
𝑇𝑇𝐻𝐻 − 𝑇𝑇𝐶𝐶

The real-world behavior of the heat pump efficiency is much less than the Carnot COP. Previous
literature introduced a Carnot factor, given here as 𝜂𝜂, which estimates the actual COP.

𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = 𝜂𝜂 ∙ 𝐶𝐶𝐶𝐶𝑃𝑃𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶

In practice, the Carnot factor (𝜂𝜂) is typically between 0.4 and 0.6 (40% to 60%). This
methodology ignores the refrigerant and compressor properties but allows the user to scan
Carnot factor ranges and evaluate whether, for a given application, a heat pump is potentially
realistic, which can lead to Method 2.

Method 2: Refrigerant and Compressor Efficiencies

Using the Python package CoolProp and the built library with compressor properties, the model
can estimate a COP based on physics. First, for all refrigerants in the process library, the model
checks whether their critical temperature (𝑇𝑇𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 ) is greater than the required high temperature, 𝑇𝑇𝐻𝐻 .
If yes, the refrigerant is added to a list that is provided to the user for further analysis. From this
list, the model selects the refrigerant with the lowest critical pressure (𝑃𝑃𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 ) for analysis because
the higher the compressor’s pressure ratio (𝑃𝑃𝑃𝑃), the less efficient the compressor’s operation.
The equation for the compressor ratio and compressor efficiency (εPR), is given here. If an error
arises from this process or the inputs are out of range, the refrigerant R1234ze(Z) (3-
Tetrafluropropoene, a refrigerant in the hydrofluoroolefin category used for some HTHPs) is
used as the default refrigerant.

𝑃𝑃2
𝑃𝑃𝑃𝑃 =
𝑃𝑃1

𝜀𝜀𝑃𝑃𝑃𝑃 = 0.95 − 0.125 × 𝑃𝑃𝑃𝑃

In some literature, a volume ratio is also given; however, the efficiency of the compressor per the
volume ratio heavily depends on the characterization of the compressor, and no general formulas
were found. Rather, each manufacturer provides tables or characterization curves. After initial
testing, the volume ratio compressor efficiency (𝜀𝜀𝑉𝑉𝑉𝑉 ) is assigned a default value of 0.75. This can
be altered by the user and is not suggested as an exact value. Rather, this is a value whose initial
estimates demonstrated reasonable COP calculations, but this should be evaluated for specific
use cases and compressor types.

Once the refrigerant is chosen, the thermodynamic enthalpy is calculated using the CoolProp
package for all points in the heat pump cycle. The COP is then calculated as:

10
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 𝑞𝑞ℎ 𝜀𝜀𝑉𝑉𝑉𝑉 𝜀𝜀𝑃𝑃𝑃𝑃 (ℎ3 − ℎ2 )
𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = =
𝑤𝑤𝑐𝑐 (ℎ2 − ℎ1 )

From either Method 1 or Method 2, an estimated “actual” COP is calculated and used for the
remainder of the analysis.

2.4.2 Single-Stage with Internal Heat Exchanger

Another configuration often used to increase the operating performance of the heat pump is the
internal heat exchanger, shown in Figure 6. The goal is to superheat the saturated vapor at the
evaporator outlet (point 1 to 1′) and undercool the saturated liquid at the condenser outlet (point
3 to 3′). Figure 6 also shows the effect of the internal heat exchanger temperature difference on
the cycle COP, highlighting that this configuration does not always increase the COP, depending
mostly on the working fluid phase-diagram shape. From a calculation perspective, because the
two sides of the internal heat exchanger have the same composition and flow rate, they will also
incur the same temperature difference; the only constraint is that the cold outlet stream
temperature must be less than the hot inlet by an approach temperature (typically 5°C). (The
same is true between the hot outlet and the cold inlet.) Figure 6 shows the COP calculation for
this configuration, which uses the modified Point 1 enthalpy (ℎ1′ ). The COP uses the unmodified
condenser outlet enthalpy (ℎ3 ) rather than the modified enthalpy (ℎ3 ,) because only the
condenser delivers heat to the heat sink.

𝑞𝑞ℎ 𝜀𝜀𝑉𝑉𝑉𝑉 𝜀𝜀𝑃𝑃𝑃𝑃 (ℎ3 − ℎ2 )

𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = =
𝑤𝑤𝑐𝑐 (ℎ2 − ℎ1′ )

Figure 6. Single-stage cycle with internal heat exchanger

11
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
2.4.3 Single-Stage with Supercritical Fluid
Another strategy is to employ working fluids that operate at a supercritical state in the condenser,
as shown in Figure 7. The main difference in this case is that the “condenser” equipment
becomes a regular heat exchanger with a temperature slope on the working fluid side (as opposed
to a mostly constant temperature process for regular cycles). This might be advantageous when
the sink-side fluid also requires a steep temperature slope (e.g., heating water from ambient
temperature to 90°C). Because the fluid is supercritical at the condenser operating conditions, it
will typically avoid very high temperatures at the compressor outlet and allow very low
temperatures in the evaporator, both of which could work favorably toward increasing the
cycle’s efficiency. A key change for this configuration is that the usual specification of the
condenser outlet as a saturated liquid is not present anymore (because it is at a supercritical
state), which also allows the condenser outlet temperature to not be linked to the sink outlet
(𝑇𝑇ℎ,𝑜𝑜𝑜𝑜𝑜𝑜 ) but to its inlet (𝑇𝑇ℎ,𝑖𝑖𝑖𝑖 ), again taking advantage of a steep slope on the sink side. In
addition, the compressor pressure ratio becomes a design variable, as shown in Figure 7. The
COP calculation for this configuration does not change from the single-stage case.

Figure 7. Supercritical cycle thermodynamic diagram and effect of pressure ratio on COP

2.4.4 Two-Stage Configurations

Two-stage cycles are usually employed for large temperature lifts (e.g., 150°C) to improve the
operational performance (COP) and to allow a feasible pressure ratio for the compressors.
Several configurations are possible; some of the most common employ equipment such as heat
exchangers, splitters, mixers, and two-phase flash separators (Cengel and Boles 2014). All these
configurations employ the same equation for the cycle COP that can also be calculated on a per
time basis:

𝑞𝑞̇ ℎ 𝑞𝑞ℎ 𝑚𝑚̇𝑐𝑐 𝑞𝑞ℎ

𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = = =
𝑤𝑤̇ℎ 𝑤𝑤𝑐𝑐1 𝑚𝑚̇𝑒𝑒 + 𝑤𝑤𝑐𝑐2 𝑚𝑚̇𝑐𝑐 𝑤𝑤 𝑚𝑚̇𝑒𝑒 + 𝑤𝑤
𝑐𝑐1 𝑚𝑚̇ 𝑐𝑐2
𝑐𝑐

where the heat and work per mass (𝑞𝑞ℎ , 𝑤𝑤𝑐𝑐1 , and 𝑤𝑤𝑐𝑐2 ) are given by the difference in enthalpy
𝑚𝑚̇
between various points in the cycle, and the mass flow rate ratio (𝑚𝑚̇𝑒𝑒) is derived from a mass or
𝑐𝑐
energy balance around the equipment linking the two stages (i.e., heat exchanger, flash tank,
mixer). The following four configurations present the corresponding COP calculations and flow
sheets shown in Figure 8:

12
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 𝑞𝑞ℎ (ℎ6 −ℎ7 )
Two-stage cascade: 𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = 𝑚𝑚̇ = (ℎ2 −ℎ3 )
𝑤𝑤𝑐𝑐1 ̇ 𝑒𝑒 +𝑤𝑤𝑐𝑐2 (ℎ2 −ℎ1 )
(ℎ8 −ℎ5 )
+(ℎ6 −ℎ5 )
𝑚𝑚𝑐𝑐

𝑞𝑞ℎ (ℎ4 −ℎ5 )

Two-stage with splitter economizer: 𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = 𝑚𝑚̇𝑒𝑒 = (ℎ −ℎ )
𝑤𝑤𝑐𝑐1 ̇ +𝑤𝑤𝑐𝑐2 (ℎ2 −ℎ1 ) 3 6 +(ℎ4 −ℎ3 )
𝑚𝑚𝑐𝑐 (ℎ2 −ℎ7 )

𝑞𝑞ℎ (ℎ4 −ℎ5 )

Two-stage with flash economizer: 𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = 𝑚𝑚̇𝑒𝑒 = (ℎ
𝑤𝑤𝑐𝑐1 ̇ +𝑤𝑤𝑐𝑐2 2 −ℎ1 )(1−β6 )+(ℎ4 −ℎ3 )
𝑚𝑚𝑐𝑐

where β6 is the quality (vapor fraction) of Point 6.

𝑞𝑞ℎ (ℎ4 −ℎ5 )
Two-stage with flash intercooler: 𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 = 𝑚𝑚̇ = (ℎ −ℎ )
𝑤𝑤𝑐𝑐1 ̇ 𝑒𝑒 +𝑤𝑤𝑐𝑐2 (ℎ2 −ℎ1 ) 6 3 +(ℎ4 −ℎ3 )
(ℎ7 −ℎ2 )
𝑚𝑚𝑐𝑐

Figure 8. Two-stage configurations (left to right, from top left): cascade, splitter economizer, flash
economizer, and flash intercooler

2.4.5 Mass and Energy Balance

The mass and energy balance portion of the model returns three arrays: an 8,760-length array for
the minimum mass flow of the hot process stream, the cold process stream, and the electrical
input needed for the heat pump. These arrays are 8,760 entries long to provide an hourly
operating profile for 1 year. The pumping power required by these flows is nontrivial in the
design of the system piping and the sizing of the heat exchangers, but these are not considered in

13
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
this model. The power draw by the heat pump compressor is the major cost driver; hence, the
authors provide the electrical draw for the economic calculations. Additionally, a heat pump
pulls thermal energy from the waste/ambient stream. The mass and energy balance also return
the final temperature of the cold-side process stream. This can be used to understand the
potential value of simultaneously providing cooling load in an industrial setting, though no
monetization mechanism is currently implemented.

To calculate the heat exchanger boundary conditions for both the hot and cold working fluids,
the specific heat (𝑐𝑐𝑝𝑝 ) and the heat requirement of the industrial process (𝑄𝑄 in terms of Btu/h) are
used. The per-second notation for the process heat is given with the dot superscript (𝑄𝑄̇ ) 3:

𝑄𝑄
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑒𝑒𝑠𝑠𝑠𝑠 ℎ𝑒𝑒𝑒𝑒𝑒𝑒 = = 𝑄𝑄̇ = 𝑚𝑚̇ ∙ ∆𝑇𝑇 ∙ 𝑐𝑐𝑝𝑝
3600

𝑘𝑘𝑘𝑘 𝑄𝑄̇
𝑚𝑚̇ � � =
𝑠𝑠 ∆𝑇𝑇 ∙ 𝑐𝑐𝑝𝑝

This calculation is the same for both the hot and cold streams, except the hot side uses a
specified minimum allowable temperature. (For example, for pasteurization, the minimum
temperature is specified as 75°C, assuming that below this temperature, pasteurization is no
longer effective.) In contrast, the cold fluid accepts a constant mass flow rate as an input and
returns the outlet cold temperature of the heat exchanger that meets this heat removal rate.

To calculate the power draw (𝑃𝑃𝑖𝑖𝑖𝑖 ) of the heat pump, the estimated actual COP is used along with
the 8,760 required heat rate. After converting from Btu/h to kWth, the sum of the 8,760 array is
taken as the annual electricity drawn by the heat pump.

𝑄𝑄
𝑃𝑃𝑖𝑖𝑖𝑖 =
𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎

Using this annual power consumption, both an average power consumption (𝑃𝑃�𝑖𝑖𝑖𝑖 ) and a capacity
factor for the heat pump can be calculated.

2.4.6 Heat Pump Lifetime Costs and Levelized Cost of Heat

The heat pump economics include the capital, operating, and energy costs of the heat pump.

Based on the maximum thermal demand, the capital cost (𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐 𝑐𝑐𝑐𝑐𝑐𝑐𝑡𝑡𝐻𝐻𝐻𝐻 ) is calculated using a
per-kilowatt user input overnight capital cost (𝑂𝑂𝑂𝑂𝐶𝐶𝑘𝑘𝑘𝑘 ) multiplied by the maximum electricity
capacity (𝐶𝐶𝐶𝐶𝑝𝑝𝐻𝐻𝐻𝐻 ).This is assumed to be an overnight capital cost because no financing is
evaluated in the current model version. For the sum, this variable is also assigned to 𝐶𝐶0 .

3
This model requires that 𝑄𝑄ℎ = 𝑄𝑄, which is used as a boundary condition for the heat pump, whereas 𝑄𝑄̇ℎ ≠ 𝑄𝑄̇
because the mass flow between the refrigerant and the process stream are not the same, only the energy flow
between the two.

14
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 𝑄𝑄
𝐶𝐶𝐶𝐶𝑝𝑝𝐻𝐻𝐻𝐻 = max � �
𝐶𝐶𝐶𝐶𝑃𝑃𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎

$
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑡𝑡𝐻𝐻𝐻𝐻 = 𝑂𝑂𝑂𝑂𝐶𝐶𝑘𝑘𝑘𝑘 � � × 𝐶𝐶𝐶𝐶𝑝𝑝𝐻𝐻𝐻𝐻
𝑘𝑘𝑘𝑘

The annual operating and energy costs of the heat pump for a single year (𝐶𝐶𝑖𝑖𝐻𝐻𝐻𝐻 ) are a function of
the fixed operation-and-maintenance (FO&M) cost, variable operation-and-maintenance
(VO&M) cost, and annual energy cost (AEC) of the heat pump. For the electricity prices, the
annual energy costs are given in an 8,760 array to match a utility rate (UR).

$ $ $
𝐶𝐶𝑖𝑖𝐻𝐻𝐻𝐻 = 𝐶𝐶𝐶𝐶𝑝𝑝𝐻𝐻𝐻𝐻 × 𝐹𝐹𝐹𝐹&𝑀𝑀 � � + 𝐴𝐴𝐴𝐴𝐴𝐴(𝑘𝑘𝑘𝑘ℎ) × 𝑉𝑉𝑉𝑉&𝑀𝑀 � � + 𝐴𝐴𝐴𝐴𝐴𝐴(𝑘𝑘𝑘𝑘ℎ) × 𝑈𝑈𝑈𝑈 � �
𝑘𝑘𝑘𝑘 𝑘𝑘𝑘𝑘ℎ 𝑘𝑘𝑘𝑘ℎ

Using this 1-year operating cost, along with the capital costs and a discount rate, the life-cycle
costs and lifetime LCOH can be calculated.
𝑁𝑁
𝐶𝐶𝑖𝑖𝐻𝐻𝐻𝐻
𝐿𝐿𝐿𝐿𝐿𝐿 = 𝐶𝐶0 + �
(1 + 𝑑𝑑)𝑖𝑖
𝑖𝑖=1

𝐿𝐿𝐿𝐿𝐿𝐿
𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 =
𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌 1 𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 × 𝑁𝑁

2.4.7 Gas Boiler

The gas model economics are similar to the heat pump economics, but there is no COP and an
efficiency of the gas boiler (𝜀𝜀𝑔𝑔 ) that converts the input energy of natural gas to meet the thermal
energy demand. The model can accept a Boolean flag that either the capital cost of the gas boiler
either to zero (suggesting an existing boiler) or to a per-unit base sized against the maximum
process heat requirement similar to the heat pump.

$ max(𝑄𝑄)
𝐶𝐶𝐶𝐶𝐶𝐶𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑐𝑐𝑐𝑐𝑐𝑐𝑡𝑡𝐺𝐺𝐺𝐺𝐺𝐺 = 𝑂𝑂𝑂𝑂𝐶𝐶𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 � �×
𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀/ℎ𝑟𝑟 𝜀𝜀𝑔𝑔

$ $
𝐶𝐶𝑖𝑖𝐺𝐺𝐺𝐺𝐺𝐺 = 𝐶𝐶𝐶𝐶𝑝𝑝𝐺𝐺𝐺𝐺𝐺𝐺 × 𝐹𝐹𝐹𝐹&𝑀𝑀 � � + 𝐸𝐸𝐸𝐸(𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀) × 𝑉𝑉𝑉𝑉&𝑀𝑀 � �
𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
ℎ𝑟𝑟
$
+ 𝐴𝐴𝐴𝐴𝐴𝐴 (𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀) × 𝑈𝑈𝑈𝑈 � �
𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
𝑁𝑁
𝐶𝐶𝑖𝑖𝐺𝐺𝐺𝐺𝐺𝐺
𝐿𝐿𝐿𝐿𝐿𝐿 = 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑡𝑡𝐺𝐺𝐺𝐺𝐺𝐺 + �
(1 + 𝑑𝑑)𝑖𝑖
𝑖𝑖=1

15
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 𝐿𝐿𝐿𝐿𝐿𝐿
𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 =
𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌 1 𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 × 𝑁𝑁

2.4.8 Electro-Resistive Heater

The electro-resistive heater economics are similar to the gas model, but fuel costs are replaced
with electricity costs. The electro-resistive heater is assigned an efficiency (𝜀𝜀𝐸𝐸𝐸𝐸 ) that converts the
input energy of electricity to meet the thermal energy demand. The electricity price includes both
energy and demand charge costs.

$ max(𝑄𝑄)
𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑡𝑡𝐸𝐸𝐸𝐸 = 𝑂𝑂𝑂𝑂𝐶𝐶𝑘𝑘𝑘𝑘 � �×
𝑘𝑘𝑘𝑘 𝜀𝜀𝐸𝐸𝐸𝐸

$ $ $
𝐶𝐶𝑖𝑖𝐸𝐸𝐸𝐸 = 𝐶𝐶𝐶𝐶𝑝𝑝𝐸𝐸𝐸𝐸 × 𝐹𝐹𝐹𝐹&𝑀𝑀 � � + 𝐴𝐴𝐴𝐴𝐴𝐴(𝑘𝑘𝑘𝑘ℎ) × 𝑉𝑉𝑉𝑉&𝑀𝑀 � � + 𝐴𝐴𝐴𝐴𝐴𝐴(𝑘𝑘𝑘𝑘ℎ) × 𝑈𝑈𝑈𝑈 � �
𝑘𝑘𝑘𝑘 𝑘𝑘𝑘𝑘ℎ 𝑘𝑘𝑘𝑘ℎ
𝑁𝑁
𝐶𝐶𝑖𝑖𝐸𝐸𝐸𝐸
𝐿𝐿𝐿𝐿𝐿𝐿 = 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑡𝑡𝐸𝐸𝐸𝐸 + �
(1 + 𝑑𝑑)𝑖𝑖
𝑖𝑖=1

𝐿𝐿𝐿𝐿𝐿𝐿
𝐿𝐿𝐿𝐿𝐿𝐿𝐿𝐿 =
𝑌𝑌𝑌𝑌𝑌𝑌𝑌𝑌 1 𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 × 𝑁𝑁

2.5 Model Economics

The final component in the model is an economics model. The economics model compares the
performance of any two systems. A system is either a stand-alone heat pump, gas, or electro-
resistive heater, or it can be a hybrid combination of several of them. The economics module
includes a subroutine to combine any two components into a system, or this combination can be
done multiple times.

Given System 1 and System 2, the economics module calculates an NPV, an IRR, and a PBP.
The cash flow is assumed to be the difference between the two systems’ capital and annual
operating costs.

𝐶𝐶0 = 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑡𝑡𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 1 − 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑡𝑡𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 2

𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 1 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 2
𝐶𝐶𝑖𝑖 = 𝐶𝐶𝑖𝑖 − 𝐶𝐶𝑖𝑖
𝑁𝑁
𝐶𝐶𝑖𝑖
𝑁𝑁𝑁𝑁𝑁𝑁 = 𝐶𝐶0 + �
(1 + 𝑑𝑑)𝑖𝑖
𝑖𝑖=1

The IRR is found by setting the NPV = 0 and solving for the discount rate (d).

𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 1 − 𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶 𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑐𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 2

𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝𝑝 =
�𝐿𝐿𝐿𝐿𝐿𝐿𝐻𝐻𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 2 − 𝐿𝐿𝐿𝐿𝐿𝐿𝐻𝐻𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 1 � × 𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴𝐴 𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑𝑑 (𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀)

16
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
3 Python Example Inputs
The HTHP model comes with preloaded default values to make it easy to build the model. This
section outlines the default values and how to load them, how to adjust them, and how to run a
full economic analysis.

3.1 Default Input Values

The default input values were taken from a basic industrial process that required an input
temperature of 120°C (generally used for steam generation in food and beverage applications).
Capital costs were taken from the literature or estimated from manufacturing partners. The
authors provide the data as realistic inputs for this report, but they would change drastically by
application and should not be used as standards. Table 3 shows these values.

17
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Table 3. Default Input Values

Electro-Resistive
Parameter Heat Pump Inputs Gas Inputs
Heat Inputs
Efficiency inputs
Carnot efficiency 0.5 NA NA
Compressor efficiency 0.7 NA NA
System efficiency NA 0.8 0.92
Refrigerant R1234ze(Z) NA NA
Working fluid
Heat source temperature
50°C NA NA
available
Hot fluid inlet temperature 90°C 90°C 90°C
Hot fluid inlet pressure 1 atm 1 atm 1 atm
Hot fluid outlet
120°C 120°C 120°C
temperature (needed)
Hot fluid outlet pressure
1.1 atm 1.1 atm 1.1 atm
(needed)
Heat exchanger buffer (on
both hot and cold sides)
Also known as heat 5°C NA NA
exchanger approach
temperature
Process heat requirement 1 MW 1 MW 1 MW
Costs
$30/kWth
Specific capital cost $300/kWth $100/kWth
(~$9000/MMBtu/h)
Fixed O&M $11.8/kWth $50/kWth $1.0/kWth
Variable O&M $0.05/kWhth $0.01/kWhth $0.01/kWhth
Electricity energy price $0.02/kWh NA $0.02/kWh
Electricity demand charge $10/kW NA $10/kW
Gas price NA $20/MMBtu NA
System lifetime 20 yr 20 yr 20 yr
Discount rate 10% 10% 10%
Compound annual growth
2.14% 2.14% 2.14%
rate
Emissions
Emissions factor NA 120,000 lb/MMSCF NA
Emissions energy
NA 9.804e-4 MMSCF/MMBtu NA
intensity
Emissions price NA $0.0/t NA

18
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
3.2 Running a Case
Included in the GitHub repository is a simple example. For the default inputs, each technology is
self-sufficient and can be run as is. The only major step for the programmer is to import and
initialize the appropriate models. From the GitHub parent directory, the following code can be
run to produce an initial case of the heat pump model. Case defaults can be overwritten using
new values following the object initialization.
import numpy as np
from utilities.unit_defs import Q_
from libraries import *
from refrigerant_properties import *
from heat_pump_model import *
from electric_model import *
from gas_model import *
from cashflow_model import *

hp_test = heat_pump()
hp_test.construct_yaml_input_quantities('heat_pump_model_inputs.yml')
hp_test.run_all('hp_test')

heater_test = electric_heater()
heater_test.construct_yaml_input_quantities('electric_model_inputs.yml')
heater_test.run_all('electric_test')

gas_test = gas_heater()
gas_test.construct_yaml_input_quantities('gas_model_inputs.yml')
gas_test.run_all('gas_test')

gas_dict = object_to_dict(gas_test)
hp_dict = object_to_dict(hp_test)

calculate_cash_flow(gas_dict, hp_dict, 20, 0.10)

19
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
4 Case Studies of Technical and Economic Potential
of Heat Pumps in the United States
The model described in this report has been used in several published case studies (Cox, Belding,
and Lowder 2022; Cox et al. 2022). Two additional case studies are presented here. The first is
the evaluation of a steam generating heat pump based on a commercial model and its application
to wort boiling for beer brewing. The second focuses on the economics of providing
simultaneous heating and cooling for pasteurization and yogurt making.

4.1 Case Study 1: Beer Brewing

The U.S. beverage industry has a high rate of energy consumption, of which a very large fraction
is associated with heating and cooling (EPA 2022). Breweries, in particular, employ intensive
heating and cooling throughout the entire production process; it is estimated that 70% of the total
energy consumed is linked to the thermal energy that is used to generate hot water and steam
(DOE 2021). Evaluations of renewable heating technologies for breweries have typically focused
on solar hot water generation and biofuels (EPA 2022), and less attention has been paid to steam
generation. Coronado Brewing Company is a medium-size beer brewery headquartered in San
Diego County, California. The brewery performs all the major steps in the beer production
process, including mashing, boiling, fermenting, finishing, and packaging. NREL researchers
engaged with plant engineers and gathered data on energy costs (e.g., electricity tariff structure
and prices, natural gas prices) as well as profiles of plant energy consumption and production
levels.

The present case study focuses on the wort boiling step, which is a major energy consumer,
typically accounting for 35%–45% of the total thermal energy consumption during brewing
(Stewart et al. 2018). This step requires steam at 115℃ and is typically performed in a kettle
using a steam-driven external heating jacket, powered by fossil-fuel-based systems, such as
natural gas boilers. Because of their ability to harness waste heat and electrify heat generation,
HTHPs can play a key role by integrating with existing energy recovery systems and
decarbonizing steam generation.

4.1.1 System Description

The general process of beer brewing is presented in various sources (e.g., Stewart et al. (2018))
and includes the main steps of malting, mashing/lautering, wort boiling, cooling/fermenting, and
conditioning/filtering/ packaging, as shown in Figure 9. Commercial breweries like Coronado
Brewing Company typically purchase finished malt from separately operated malt houses,
avoiding the need for on-site malting. Mashing converts the malt starches into fermentable
sugars by mixing the malt with hot water in a vessel called the mash tun at temperatures near
60°C to 70°C. The sugary liquid product of this step is called wort, which is extracted from the
mash tun by a process called lautering. The wort is then mixed with hops for flavor and boiled in
a copper kettle at high temperatures (100°C) for 45–90 minutes. The boiled wort mixture is
cooled to the fermentation temperature (approximately 10°C to 20°C, depending on the type of
beer), and then pitched with yeast to perform the fermentation process. Final steps include aging
the beer product (cold conditioning/storage), filtering for flavor, carbonation, and packaging.

20
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Figure 9. Beer brewing process and associated temperature requirements
Figure by NREL

Figure 10 shows the major sources of energy consumption in this process. Approximately 50%
of the heating (largely provided by natural gas) is used by the brewhouse, where the wort boiling
process occurs, and this corresponds to 80%–95% of that percentage (i.e., approximately 35%–
45% of the total heating).

HEAT ELECTRICITY NATURAL GAS

CONSUMPTION
Refrigeration Brewhouse
Brewhouse
Fermentation and Storage
Packaging Packaging
Filtration Other Utilities
Filling Bottles Compressed Air Space Heating
Filling Barrels/Kegs
Brewhouse
Administrative Department
Lighting
6% Boiler House
9%

6%5% 10%
47% 7%
35% 20%
26% 10% 45%

12%
3%
9%
25% 25%

Figure 10. Heat consumption by source (left) and electricity and natural gas consumption by
source (right)
Sources: Eßlinger 2009 (left) and EPA 2022 (right). Figure by NREL with data collected from industry interviews

The wort boiling step was chosen as the focus of our study because of its high impact in the
energy consumption of the brewery. In particular, the high temperature required for boiling the
wort mixture provides a good example for evaluating the applicability of HTHPs.

21
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
4.1.2 Model
An HTHP model was applied based on the most commonly used commercial heat pump unit for
steam generation, the Kobelko steam generating heat pump. Information about the specific
model can be found in Kaida et al. (2015), Watanabe et al. (2014), and Kobelco (n.d.). Table 4
shows the specifications for the two available steam generating heat pump models, and Figure 11
shows the general thermodynamic operating points of the HTHP cycle, adapted from Watanabe
et al. (2014).

Table 4. Specifications of Kobelco Steam Generating Heat Pumps (SGH120 and SGH165)
Source: Watanabe et al. 2014

Model SGH120 SGH165

Steam pressure
0.1 0.6
(MPa gauge)
Waste heat temperature
65 70
(°C)
Steam flow rate (t/h) 0.51 0.89
Heating power (kWth) 370 660
COP 3.5 2.5
Mixture of R245fa
Refrigerant R245fa
and R134a
Two-stage Single-stage
Refrigerant compressor
twin-screw twin-screw
Number of steam
0 1
compressors

22
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Figure 11. Two-stage heat pump cycle using R245fa and splitter/mixer between stages
SGH120 generates steam at 120°C (at approximately 1 atm) with a COP of 3.5, assuming a
source temperature of 65°C. It uses a two-stage twin-screw compressor and R245fa as the
working fluid, which has a critical temperature above 150℃ and is thus a suitable candidate for
steam production. Based on the provided thermodynamic diagram and equipment operating
conditions, it was assumed that the configuration is a two-stage cycle with intermediate splitter
economizer (Point 6 in Figure 11) and mixer (Point 3). The split fraction on Point 6 is equal to
the amount required to bring Point 2 to its saturated vapor state (Point 3). The flow sheet of this
configuration is presented in Figure 8 (top right), and its COP is given as:

𝑞𝑞̇ ℎ 𝑞𝑞ℎ 𝑚𝑚̇𝑐𝑐 𝑞𝑞ℎ (ℎ4 − ℎ5 )

𝐶𝐶𝐶𝐶𝐶𝐶 = = = =
𝑤𝑤̇𝑐𝑐 𝑤𝑤𝑐𝑐1 𝑚𝑚̇𝑒𝑒 + 𝑤𝑤𝑐𝑐2 𝑚𝑚̇𝑐𝑐 𝑤𝑤 𝑚𝑚̇𝑒𝑒 + 𝑤𝑤 (ℎ − ℎ6 )
𝑐𝑐1 𝑚𝑚̇ 𝑐𝑐2 (ℎ2 − ℎ1 ) 3 + (ℎ4 − ℎ3 )
𝑐𝑐 (ℎ2 − ℎ7 )

In this equation, the subscripts refer to the points in the thermodynamic cycle shown in Figure
11. Note that other cycle configurations are suggested in the literature. For example, the Kobelco
equipment sheet (Kobelco n.d.) suggests the existence of an internal heat exchanger in addition
to the splitter economizer; thus, future work should analyze this alternative configuration.

The HTHP model was evaluated for varying temperature lifts to assess the impact of the source
temperature on the equipment performance, assuming a fixed sink temperature of 120℃. Figure
12 presents the COP curve predicted by the model as a function of the lift temperature, with

23
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
representative operating conditions as points. Table 5 shows the experimental data and the model
outputs. The compressor isentropic efficiency was adjusted to match the model results to the
available data and was found to be 86%. This value is in the higher range for compressor
efficiencies and might be overcompensating for other differences in the cycle, such as condenser
outlet subcooling degree, cycle configuration (e.g., internal heat exchanger), and
operation/control strategy.

Figure 12. HTHP model COP compared to temperature lift (sink temperature fixed at 120°)

Table 5. Experimental Results for SGH120

Source: Data from Kobelco n.d.

Temperature (°C) HTHP

Kobelco
Source Source Sink Sink Model
Lift COP COP
Inlet Outlet Inlet Outlet
65 60 20 120 60 3.5 3.49
55 55 20 120 65 3.1 3.18
45 45 20 120 75 2.7 2.69
35 35 20 120 85 2.4 2.3

4.1.3 Inputs
The heat requirement for both the preheating and the boiling steps was calculated as follows. It
was assumed that a volume of 42 barrels (1 barrel of beer is 31 gallons) of wort-water mixture
(with added hops) is brought to a boil from an initial temperature of 70℃ (from the mashing
process), performed during the course of 1 hour. The boil itself is assumed to evaporate a volume
of 1.5 to 2 barrels of water, also during the course of 1 hour, resulting in a total of 2 hours for the
wort boiling process, including preheating. The sensible heat requirement for the preheating
stage and the latent heat for the evaporation stage are calculated as follows:

• 𝑄𝑄 𝑝𝑝𝑝𝑝𝑝𝑝ℎ𝑒𝑒𝑒𝑒𝑒𝑒 = 𝑀𝑀𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤,𝑝𝑝𝑝𝑝𝑝𝑝ℎ𝑒𝑒𝑒𝑒𝑒𝑒 𝐶𝐶𝑝𝑝𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤 Δ𝑇𝑇 𝑝𝑝𝑝𝑝𝑝𝑝ℎ𝑒𝑒𝑒𝑒𝑒𝑒 =

gals 𝑘𝑘𝑘𝑘 𝑘𝑘𝑘𝑘 1 𝑘𝑘𝑘𝑘ℎ
42[barrels]31 �𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏� 3.79 �𝑔𝑔𝑔𝑔𝑔𝑔� 4.18 �𝑘𝑘𝑘𝑘 𝐾𝐾� (100 − 75)[𝐾𝐾] 3600 � 𝑘𝑘𝑘𝑘
� = 143.9 𝑘𝑘𝑘𝑘ℎ

24
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 gals 𝑘𝑘𝑘𝑘 𝑘𝑘𝑘𝑘 1 𝑘𝑘𝑘𝑘ℎ
• 𝑄𝑄 𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 = 𝑀𝑀𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤𝑤,𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏 Δ𝐻𝐻 𝑣𝑣𝑣𝑣𝑣𝑣 = 2 [barrels] 31 �𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏𝑏� 3 .79 �𝑔𝑔𝑔𝑔𝑔𝑔� 2215.48 �𝑘𝑘𝑘𝑘� 3600 � 𝑘𝑘𝑘𝑘
�=
144.6 𝑘𝑘𝑘𝑘ℎ (𝑡𝑡ℎ𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒) (or 108.5 kWhth for 1.5 barrels)

This calculation does not account for the thermal losses to the ambient during both the
preheating and boiling stages. It was assumed that the thermal energy consumed during the
boiling stage consists of the maximum of these two values, approximately 144 kWth, because it is
typical to operate the steam generation equipment at a constant level; thus, the electricity
consumed by the heat pump will range from 72 kW (COP = 2) to 36 kW (COP = 4). Using utility
bills from the brewery, as shown in Figure 13, the monthly electricity consumption was
estimated to be approximately 40,000 kWh, equivalent to an hourly consumption of 83 kW
(assuming 24-hour operation, 20 weekdays per month); thus, the heat pump consumes
approximately one-half to one times the whole brewery electricity consumption, which agrees
with the energy consumption statistics previously provided—70% of the total energy
consumption for the brewery comes from heating and 30% comes from electricity; and the wort
boiling process consumes approximately 45% of heating energy, or approximately 30% of the
total energy, which is the same magnitude as the brewery electricity consumption. Also using the
utility bills, the authors identified the brewery monthly peak demand of electricity to be between
140 kW–180 kW; thus, the HTHP would contribute to a 25%–50% greater peak demand charge
(assuming the peak loads fall during the same time and are additive).

Figure 13. Brewery monthly energy consumption for 1 year from an SDG&E utility bill. Peak
demand data are available only for the month of June 2022 and for the yearly maximum.
Source: SDG&E 2023

To determine the capacity or utilization factor of the equipment, it was assumed that the
operation is for 2 hours (1 hour preheating plus 1 hour boiling), followed by a 4-hour rest period,
4 times per day, during weekdays only. Figure 14 shows the operation during the first week of
2021. (January 2021 started on a Friday, so the following 2 days, during the weekend, did not
have any thermal load.)

25
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Figure 14. Wort boiling thermal load profile (operation schedule of the heat pump)
Figure by NREL with data from sources (SDG&E 2023)

The brewery participates in the electricity tariff “AL-TOU2” from SDG&E (SDG&E 2023), for
medium and large commercial and industrial services, which comprises a time-of-use (TOU)
component and a demand charge component. The TOU price profile, shown in Figure 15,
includes three modes—on-peak, off-peak, and super off-peak—which depend on the time of day
and of the year (summer versus winter). Each graph in Figure 15 shows the on-peak, off-peak
and super off-peak pricing for weekday (left), weekend (middle), and combined (right).

Figure 15. TOU electricity prices for AL-TOU2 from SDG&E

Figure by NREL with data from sources (SDG&E 2023)

The demand charge comprises the following components: a non-coincident demand charge
($18.63/kW) based on the maximum monthly demand, a TOU demand charge ($34.2/kW) based
on the maximum monthly demand measured during the on-peak and off-peak time periods, and
an on-peak period demand charge ($4.26/kW during summer and $1.12/kW during winter) based
on the maximum demand during the on-peak period. These components are additive, meaning
that a maximum demand during summer on-peak hours would incur a cost of 18.63 + 34.2 +
4.26 = $57.09/kW.

The authors assumed that the natural gas price is fixed for each month of the year and calculated
these monthly values using plant data with variable and fixed components (see Figure 16).

26
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Figure 16. Monthly natural gas prices for 1 year
Table 6 presents a summary of the presented and additional inputs.

Table 6. Input Summary for Beer Brewing Case Study

Preheating requirement 144 kWth

Boiling requirement 93.30 kWth

HTHP utilization factor 2 h every 6 h, weekdays only

Electricity TOU energy price $0.06/kWh to $0.13/kWh

Electricity demand charges $54/kW to $57/kW

Electricity fixed annual cost $7,850/yr

Natural gas monthly cost $7.83/MMBtu to $12.49/MMBtu

Natural gas fixed annual cost $6,329/yr

Boiler efficiency 80%

Sink outlet temperature (steam) 115°C (at 10 psi)

Source outlet temperature (wastewater) 70°C

Number of years for LCC calculation 25

Interest rate 5%

HTHP capital cost $800/kWth

Boiler capital cost $100/kWth

4.1.4 Results
Figure 17 presents the operating costs for both the electric heat pump and the natural gas boilers.
The demand charges represent the majority of the costs of the heat pump operation, near 60%;
the TOU energy charges represent 20%; and the fixed annual costs represent the rest. For the
boiler, variable costs (i.e., that scale with the natural gas consumption) represent roughly 70% of
the total. Solely based on the operating costs (fixed plus variable components), the authors
estimated the annual cost of electricity to be $28,670/yr and natural gas to be $13,039/yr.

27
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Figure 17. Operating costs of the variable and fixed components of the electric heat pump and
natural gas boilers
Next, the LCCs were calculated for both the electric heat pump and the natural gas boilers, first
considering the capital cost for the boiler (i.e., a grassroots design). Figure 18 presents the
fraction of the capital and operating costs on the LCCs. For the HTHP, 22% of its LCC is
associated with the capital cost, which is eight times greater than that of the boiler. The HTHP
LCC was $550,940, compared to $208,796 for the boiler, resulting in an unfavorable NPV of -
$342,143.

Figure 18. Impact of capital and operating costs on the LCC of the electric heat pump and natural
gas boilers
If a boiler already exists and does not need to be replaced (i.e., no capital cost), the natural gas
boiler LCC becomes $192,964, resulting in an NPV of -$357,975.

Due to the unfavorable scenario for the implementation of HTHPs, a sensitivity analysis was
performed to analyze the effect of the heat pump COP and natural gas price on the project NPV.
Figure 19 presents the results, showing the HTHP LCC as a function of its COP (blue curve) and
the boiler LCC for multiple price multipliers of natural gas. Even for unrealistically high COPs
(i.e., 8), the project would not be economically feasible under the current natural gas prices. For
a heat source at 70ºC (commonly found in breweries’ hot water storage), the natural gas price
would need to be three times as high to make this project feasible.

28
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Figure 19. Effect of COP and natural gas price on equipment LCC
As a final analysis, the authors analyzed batteries as an alternative to reduce the demand charges
for the HTHP system. The NREL tool Renewable Energy Integration and Optimization®
(REopt®) 4 was used to design a battery energy storage system that minimizes the operating costs
while considering the specific electricity tariff structure. Table 7 presents the obtained results for
the business-as-usual (BAU) and battery cases. Figure 20 shows the optimal dispatch of the
battery. The sum of the “battery serving load” and the “grid serving load” is equivalent to the
BAU load. The battery project has a favorable NPV of $177,360, which would reduce the HTHP
LCC from $550,940 to $373,580 and the HTHP NPV from -$342,143 to -$164,783. Although
this is significant, the battery energy storage is not economic enough to enable HTHPs.

Table 7. Battery Energy Storage Design Results from NREL’s REopt

Battery power and capacity 25 kW, 67 kWh

Potential life cycle savings (25 yr) $177,360
Battery capital cost $45,566
Average annual energy supplied from grid 82,642 kWh (BAU), 88,389 kWh (battery)
Year 1 utility energy cost $10,153 (BAU), $10,425 (battery)
Year 1 utility demand cost $30,708 (BAU), $11,818 (battery)

4
See [Link]

29
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Figure 20. Battery energy storage optimal dispatch profile

4.2 Case Study 2: Pasteurization and Yogurt Making

4.2.1 System Description and Inputs
Commonwealth Dairy is a yogurt and dairy processing facility in Brattleboro, Vermont. NREL
researchers were introduced to the facility by Efficiency Vermont, an organization within the
Vermont State Energy Department. On average, Commonwealth Dairy processes approximately
95,000 pounds of dairy product per day. It does this through a pasteurization and processing
system that uses two propane water heaters and a glycol chiller. These systems provide hot and
cold loops of 98°C and -2°C, respectively.

Commonwealth Dairy is of high interest because it is considering replacing its chiller facilities
and could potentially replace both the propane water heaters and glycol chiller with a single
HTHP. The authors estimated that the BAU case of separately replacing the water heaters and
glycol chillers would cost approximately $750,000. In Vermont, energy efficiency has a high
value because the high cost of propane heating in Vermont in contrast to natural gas heating in
other areas of the United States makes a favorable price regime for HTHPs.

Commonwealth Dairy provided daily thermal energy demand based on its daily dairy product
production. The daily demand spanned from January to July and was therefore repeated, as
shown in Figure 21, starting in August.

30
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
 Figure 21. Commonwealth Dairy daily thermal energy demand (derived from energy bills)
Vermont does not have a significant natural gas pipeline infrastructure, so a significant number
of heat applications are met with more expensive propane. The electricity cost is comparable to
the national average (~$0.12/kWh) (EIA 2021), but the electricity rate significantly varies during
the large on-peak versus off-peak time periods ( shown in Table 8). There are both off-peak and
on-peak demand charges, but the authors chose the more expensive one because the model
cannot handle both. Table 8 shows the inputs for the potential HTHP system compared to the
propane and chiller system. This case is highly interesting because it would provide simultaneous
heating and cooling of the facility.
Table 8. Commonwealth Dairy Input Values (derived from energy bills)

Heat Pump Propane + Chiller

Cost $1,200/kWth $750,000
Electricity cost demand charge on-peak
$16.40/kW NA
(6 a.m.–11 p.m.)
Electricity cost demand charge off-peak $4.723/kW NA
Electricity cost energy charge on-peak
$0.11573 NA
(6 a.m.–11 p.m.)
Electricity charge off-peak $0.08795 NA
Propane cost NA $19.83/MMBtu

For this analysis, it was assumed that the chiller energy costs are not added to the gas system
costs. Additionally, it was assumed that the off-peak demand charges are significantly smaller
than the on-peak demand charges. Disregarding the chiller costs makes the heat pump less
competitive than the traditional system, and disregarding the off-peak demand charges makes the
heat pump more competitive; therefore, disregarding both has a leveling effect. This was
necessary because currently the model does not have a chiller component and cannot handle two
demand charges. Future work should fix this before a decision is made by Commonwealth Dairy.
Additionally, it was assumed that more hot water than cold water is always needed and that
heating would drive the simultaneous heating and cooling HTHP. To justify this assumption, a

31
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
30% higher capital cost for the HTHP was assumed to accommodate the potential need for
chilled glycol, chilled water, or ice energy storage, which could be used as a buffer. A discount
rate of 10% was used for the financial calculations, with an assumed system lifetime of 20 years.
The lifetime of 20 years was chosen rather than the default value of 25 based on site engineer
preferences.

4.2.2 Results
The Commonwealth Dairy case study shows a positive NPV of $371,225, making it potentially
economically viable for HTHP deployment. Previous work has often found that HTHPs are not
highly economic with COPs less than 2; in this case, however, due to the increased capital cost
of the combined water heater and chiller, the LCOH of the propane-driven hot water is very high
($26.60/MMBtu), meaning even a poorly performing HTHP can provide value if it is used for
simultaneous heating and cooling. Table 9 summarizes the results.

Table 9. Commonwealth Dairy Key Indicator Results

Average COP 1.7

Average power draw 468 kW
Capital cost $960,000
HTHP lifetime LCOH $24.95/MMBtu
Propane lifetime LCOH $26.60/MMBtu
NPV $371,225
Internal rate of return 0.003%

As mentioned, additional work is needed before deploying HTHPs at the dairy facility; however,
the initial economic analysis is promising and provides example inputs that can be used for the
HTHP model.

32
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
5 Discussion and Conclusions
This report presents a summary of the work related to NREL’s HTHP model. 5 The goal of this
model is to help estimate the economic potential of heat pumps in industrial applications to
demonstrate the value of adopting this technology. Heat pumps for space heating and industrial
applications have been used significantly throughout the world for space heating, especially in
regions with high energy prices, such as Europe, but they have yet to be heavily used in the
United States. This publicly available model and the case studies in this work could help spur
U.S. adoption.

Note that additional work should be performed to align the model with the real-world
performance of HTHPs based on manufacturer feedback. Outside of the United States, there are
several HTHP manufacturers (Arpagaus et al. 2018), though United States domestic
manufacturers could not be found. In this work, heat pump manufacturers shared insight into
refrigerant selection and compressor performance that impacted the final HTHP model. Building
a more robust HTHP model, however, will require additional manufacturer input, especially
because most HTHP applications are bespoke rather than off the shelf, which could hinder
adoption. Either as HTHP models are released or in partnership with their design, this work
could benefit from additional data from real-world HTHP facilities.

Many of the case studies evaluated with the HTHP model demonstrated economic
competitiveness; however, this economic competitiveness was exhibited with a low IRR, such
that it is understandable that many customers would not want to be the first movers on a novel
technology, especially if the technology has only a slight edge economically over more
established technologies, such as natural gas heaters. Efforts to reduce the capital cost or increase
the economic competitiveness, such as accounting for the cost of CO2 emissions, could help
address this issue.

During this work, experts in both heat pump manufacturing and food and beverage processing
expressed anecdotal concerns about the unproven heat pump technologies based on experience in
the United States, but they were generally open to heat pumps as a decarbonizing technology.
Several barriers cited by experts included the nascent technology, the lack of commercially
available heat pump models, either from domestic or international suppliers, the lack of off-the-
shelf HTHP solutions (e.g., many HTHP solutions would require significant piping and
integration systems to incorporate the HTHP into industrial processes), and unproven operation-
and-maintenance costs. Any additional related research that could be performed, such as
demonstration or pilot projects, could be useful in alleviating these concerns.

Heat pumps are already a cost-effective technology for many heating applications, and HTHPs
are a promising technology with the potential to reduce energy consumption, increase energy
efficiency, reduce customer energy bills, and reduce CO2 emissions (depending on the emissions
intensity of the grid or other electricity source). As their operating temperatures are pushed
higher, or as they provide simultaneous heating and cooling, additional industries open up as
potential users of HTHPs. The model described in this report is publicly available and can be

5
See [Link]

33
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
used to estimate the performance and economic competitiveness of HTHPs. Future work could
seek to couple this model to experimental demonstration and pilot project facilities to increase its
robustness. Until then, the selected case studies show the potential for HTHPs as a viable
industrial heat source in the U.S.

34
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
References
Arpagaus, Cordin, and Stefan Bertsch. 2020. “Industrial Heat Pumps in Switzerland –
Application Potentials and Case Studies”. Technical Report, Swiss Federal Office of Energy
SFOE. URL: [Link]

Arpagaus, Cordin, Frédéric Bless, Michael Uhlmann, Jürg Schiffmann, and Stefan S. Bertsch.
2018. “High Temperature Heat Pumps: Market Overview, State of the Art, Research Status,
Refrigerants, and Application Potentials.” Energy 152 (June): 985–1010.
[Link]

Cengel, Yunus, and Michael Boles. 2014. Thermodynamics: An Engineering Approach. 8th
edition. New York, NY: McGraw-Hill Education.

Cox, Jordan Tracy, Gustavo Campos, Paul Armatis, Scott Belding, Travis Lowder, and Andreas
Zourellis. 2022. “Low-Carbon District Heating: Performance Modeling of Hybrid Solar, Heat
Pump and Thermal Storage Systems for District Thermal Energy in the United States.” Presented
at the 19th International Refrigeration and Air Conditioning Conference at Purdue, July 10–14,
2022.

Cox, Jordan, Scott Belding, and Travis Lowder. 2022. “Application of a Novel Heat Pump
Model for Estimating Economic Viability and Barriers of Heat Pumps in Dairy Applications in
the United States.” Applied Energy 310 (15).

David, A., B. Vad Mathiesen, H. Averfalk, S. Werner, H. Lund. “Heat Roadmap Europe: Large-
Scale Electric Heat Pumps in District Heating Systems.” Energies 10 (2017) : 578.
[Link]

DOE Energy Efficiency and Renewable Energy. “Industrial Heat Pumps for Steam and Fuel
Savings.” Industrial Technologies Program (2014)
[Link]

DOE. “Chapter 6: Innovating Clean Energy Technologies in Advanced Manufacturing –

Technology Assessments: Process Heating.” Quadrennial Technology Review (2015).
[Link]

DOE. 2021. “Drink Up Some Energy Savings: Celebrating National Beer Day.” Accessed
November 10, 2022. [Link]
national-beer-day.

EIA. 2021. “U.S. Electricity Profile 2021.” [Link]

EPA. RHC for Breweries. (2022). Accessed on 11/10/2022. URL: [Link]

breweries

Eßlinger, Hans Michael (Editor). Handbook of Brewing: Processes, Technology, Markets, 1st
Edition. Wiley-VCH. (2009). ISBN: 978-3-527-62348-8.

35
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
Jakobs, Rainer, and Carmen Stadlander. 2020. Annex 48 Industrial Heat Pumps, Second Phase
Final Report. International Energy Agency.

Jianyong, C., H. Havtun, B. Palm. “Parametric analysis of ejector working characteristics in the
refrigeration system.” Applied Thermal Engineering 69 (2014): 130-142.
[Link]

Kaida, Takenobu, I Sakuraba, K Hashimoto, H Hasegawa. Experimental performance evaluation

of heat pump-based steam supply system. IOP Conf. Ser.: Mater. Sci. Eng. (2015). 90 012076

Kaida, Takenobu, Tsukasa Takayama, and Hiroyoshi Kataoka. 2017. “Application of Water-
Circulation Type Heat Pump to Lumber Drying System.”

KOBELCO. Steam Grow Heat Pump / SGH165 Technical Sheet. (No date). Accessed November
10, 2022. URL: [Link]
content/uploads/sites/70/2022/07/[Link]

Kosmadakis, George, Cordin Arpagaus, Panagiotis Neofytou, and Stefan Bertsch. 2020.
“Techno-Economic Analysis of High-Temperature Heat Pumps with Low-Global Warming
Potential Refrigerants for Upgrading Waste Heat up to 150 °C.” Energy Conversion and
Management 226 (December): 113488. [Link]

Laue, H.-J., K. Lassman. “Annex 35: Application of Industrial Heat Pumps – Final Report.” IEA
Heat Pump Centre HPP-AN35-1 (2014).
[Link]

Liew, P. Y., T. G. Walmsley. “Heat Pump Integration for Total Site Waste Heat Recovery.”
Chemical Engineering Transactions 52 (2016): 817-822. [Link]

Madeddu, Silvia, Falko Ueckerdt, Michaja Pehl, Juergen Peterseim, Michael Lord, Karthik Ajith
Kumar, Christoph Krüger, and Gunnar Luderer. 2020. “The CO2 Reduction Potential for the
European Industry via Direct Electrification of Heat Supply (Power-to-Heat).” Environmental
Research Letters 15 (12): 124004. [Link]

McMillan, C. A., M. Ruth. “Using facility-level emissions data to estimate the technical potential
of alternative thermal sources to meet industrial heat demand.” Applied Energy 239 (2019):
1077-1090. [Link]

Rony, R.U., H. Yang, S. Krishnan, J. Song. “Recent advances in transcritical CO2 (R744) heat
pump system: a review.” Energies 12 (2019): 457. [Link]

Ruth, M., C. McMillan, P. Kurup. “Low-Carbon Energy for Industry.” Joint Institute for
Strategic Energy Analysis Annual Meeting (2019).
[Link]

San Diego Gas & Electric Company (SDG&E). 2023. Electric Tariff Book,
Commercial/Industrial rates, SCHEDULE AL-TOU2. Accessed January 2, 2023,
[Link]

36
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.
Schlosser, F., M. Jesper, J. Vogelsang, T. G. Walmsley, C. Arpagaus, and J. Hesselbach. 2020.
“Large-Scale Heat Pumps: Applications, Performance, Economic Feasibility and Industrial
Integration.” Renewable and Sustainable Energy Reviews 133 (November): 110219.
[Link]

Stewart, Graham G., Inge Russell, Anne Anstruther (Eds). Handbook of Brewing, 3rd Ed. (2017).
CRC Press, Boca Raton. [Link]

Tveit, Tor-Martin. 2017. “Application of an Industrial Heat Pump for Steam Generation Using
District Heating as a Heat Source.”

Verdeyen, E. “Introducing the Qpinch Chemical Heat Transformer. A breakthrough innovation

to boost plant efficiency and your company’s EBITDA.” European Refining Technology
Conference (2017).

Watanabe, Choyu, Yohji Uchiyama, Satoshi Hirano, Takeshi Hikawa. Pioneering Industrial Heat
Pump Technology in Japan. 11thIEA Heat Pump Conference 2014, May 12-16, 2014, Montréal
(Québec) Canada.

Wolf, S., M. Blesl. “Model-based quantification of the contribution of industrial heat pumps to
the European climate change mitigation strategy.” ECEEE Industrial Summer Study Proceedings
(2016): 477-487.
[Link]
technology-products-and-systems/model-based-quantification-of-the-contribution-of-industrial-
heat-pumps-to-the-european-climate-change-mitigation-strategy/

Zühlsdorf, Benjamin, Jonas Kjær Jensen, Stefano Cignitti, Claus Madsen, and Brian Elmegaard.
2018. “Analysis of Temperature Glide Matching of Heat Pumps with Zeotropic Working Fluid
Mixtures for Different Temperature Glides.” Energy 153 (June): 650–60.
[Link]

37
This report is available at no cost from the National Renewable Energy Laboratory at [Link]/publications.