CHEMICAL ENGINEERING TRANSACTIONS

VOL. 81, 2020

A publication of

The Italian Association
of Chemical Engineering
Online at www.cetjournal.it

Guest Editors: Petar S. Varbanov, Qiuwang Wang, Min Zeng, Panos Seferlis, Ting Ma, Jiří J. Klemeš

ISBN 978-88-95608-79-2; ISSN 2283-9216

Design and Integration of Heat Recovery in Combination with

Solar and Biomass-based Heating in a Drying Plant

Elin Wallin, Rickard Fornell*, Oskar Räftegård, Tommy Walfridson, Jessica Benson

RISE Research Institute of Sweden, Box 857, 50115, Borås, Sweden

Rickard.Fornell@ri.se

The EU27 annually generates 90 Mt of food waste, and approximately 40 % of this waste is generated during

manufacturing. The food processing industry needs to develop improved and sustainable solutions for waste

valorisation and re-use. The project DRALOD addresses this issue since it aims at design, integration and

assessment of the performance of a heat recovery system in connection with an innovative low-temperature air

drying unit where high moisture food waste is dewatered and sold as a by-product with preserved nutritional

ingredients. The air used for drying is preheated using solar heat and a biomass boiler, and in this project the

potential benefits of integration of a heat recovery system have been investigated. Due to the impurities available

in the humid exhaust air from the dryer, the heat recovery is designed with two principal systems; a wet scrubber

condensation system and a heat pump system. Simulations using hourly meteorological data from Madrid have

been made for the total system with heat recovery, and the generated results have been used as inputs in a

techno-economic analysis in order to assess the integration of the heat recovery and how sensitive different

economic parameters and assumptions are for the results. It can be concluded from the assessment that the

prices of electricity and biofuel will have a high impact on the economic performance and design. For the

integration of the heat recovery system to be economically justifiable it is estimated, given assumptions made

in this analysis, that the electricity cost needs to be less than 5 times higher than the biofuel cost.

1. Introduction

For sustainable use of resources, it is important to minimise waste by finding innovative uses for residual

streams. In order to create value from waste streams in the food industry water often needs to be removed.

Drying and evaporation are two technologies that are often used for this purpose, but they are both energy-

intensive unit operations (Marechal and Muller, 2008). In comparison with most of the energy-intensive

industries, such as petrochemical, paper and iron and steel production, the processing of food also requires

relatively low temperature levels. As a result, the food production sites can typically integrate heat pumping or

combined heat and power solutions to upgrade waste energy (Marechal and Muller, 2008). This paper is part of

the project DRALOD, a Fast Track to Innovation (FTI) project devoted to deliver an innovative drying plant

concept with the main aim of bringing financial feasibility to the valorisation and reuse of waste from the food

processing industry. The project scope is the design, implementation and demonstration of a low-temperature

air drying technology based on the combined use of renewable energy sources (solar and biomass) for air

heating. The solar air heating technology implemented is the SolarWall patent protected technology, using

micro-perforated metal collectors that create negative pressure in an air cavity when heated by solar radiation.

A biomass heating system with high flexibility concerning type of fuels, fuel quality and mode of operation (full

and part load, dynamic operation behavior) enabling low operational costs is included as an auxiliary heat source

for the drying system. Finally, a heat recovery system adapted to the characteristics of the plant (high air mass

flow, low drying temperature, impurities in the exhaust air used as waste heat source) is developed to increase

the energy efficiency of the drying process. Compared to e.g. the systems described in (Alves-Filho, 2018) the

system assessed here is not an integrated heat pump dryer, but rather a system with a specific dryer technology

integrated with renewable heating in the form of solar and biomass, and with a heat recovery unit to increase

the total energy efficiency. The system is similar to one of the solar thermal and heat pump systems described

DOI: 10.3303/CET2081232

Paper Received: 30/04/2020; Revised: 28/05/2020; Accepted: 01/06/2020
Please cite this article as: Wallin E., Fornell R., Räftegård O., Walfridson T., Benson J., 2020, Design and Integration of Heat Recovery in
Combination with Solar and Biomass-based Heating in a Drying Plant, Chemical Engineering Transactions, 81, 1387-1392
DOI:10.3303/CET2081232

1387

in (Grubbauer et al., 2018) although the DRALOD system is open and aims at heating ambient air, also contains

a biomass boiler and operates at lower temperatures. These differences imply that there are other problems

related to system performance, e.g. the variability in input conditions (temperature, humidity) and the amount of

impurities in the heat source.

2. System description

The drying plant in this assessment is designed to dry Brewer's Spent Grains (BSG), residues from a brewery.

The BSG is expected to be sold as a product used for nutraceuticals and pharmaceutical applications, which

implies that the temperature in the drying process needs to be low in order to not deteriorate the quality of the

BSG. The drying system is divided into four sub-systems: the solar roof, the Heat Recovery Unit (HRU), the

biomass boiler and the dryer. The HRU is in turn made up of two principal systems, the wet scrubber

condensation system and the heat pump system, see Figure 1. The dryer system is specified to dry above

35,000 t/y of wet BSG from 80 % wet content to 15 % wet content with 75 °C supply air temperature. In order

to maintain the necessary driving force for drying, the air is not allowed to reach saturation, but will reach 85 %

relative humidity.

Figure 1: Overview of the drying system and the four subsystems; Solar Roof, Biomass Boiler, Dryer and HRU

(i.e. wet scrubber and Heat Pump).

From an energy perspective this is in theory a very simple system since there are only two relevant streams to

consider (since the BSG is not preheated prior to the drying plant). First the air used in the dryer needs to be

heated from ambient temperature to the specified inlet temperature of supply air in the dryer (75 °C). Then the

moist air leaving the dryer at 35 °C and 85 % humidity is either emitted to the atmosphere or, if an HRU is

included in the system, recovered in order to preheat the supply air. Two practical problems need to be

addressed in this system in order to recover the heat in the exhaust air: 1) the highly dynamic nature due to the

use of ambient air for drying and the primary energy source being solar heat, and 2) the potential impurities in

the exhaust air flow from the dryer. These practical issues have been addressed by using an operational strategy

based on the supply of a constant volumetric air flow to the drier of 180,000 m3/h. The solar heat is the first

heating source in the system, and the heat pump is assumed to heat the supply air after the solar roof. Finally,

the biomass boiler will heat the supply air up to 75 °C. Due to the impurities in the exhaust air a wet scrubber

system, connected to a water loop where heat can be delivered to the evaporator of an ammonia-based heat

pump, is included in the system. For a case without heat recovery cyclones can be used to remove the impurities.

The wet scrubber system will decrease the temperature of the waste heat with a few degrees, whilst the majority

of the latent heat of water in the exhaust air is condensed. The main issue to address in this system is to handle

the highly dynamic nature of a drying plant by using solar heating of ambient air. This will lead to a large variation

in conditions and demands a flexible operation of the system. As stated by e.g. (Schosser et al., 2019) the

proper placement of a heat pump is across the Pinch point, but in practice in a system with large variation this

might not be 100 % possible in practice. In Figure 2 two different operating conditions are shown, the upper

curves show Pinch curves for the system when the ambient temperature is low and the sun is not shining (i.e.

1388

cold stream (air) needs to be heated from a low temperature) and the lower curves where the temperature is

higher and the solar heating is active. As seen in the upper curves in Figure 2a small amount of direct Heat

Integration could in theory be possible at time points when the solar heating is not active, and the ambient

temperature is low enough (e.g. winter or night time). As seen in the lower curves in Figure 2 there is no

possibility for direct integration when solar heating is included in the system, and this is also the case if the

ambient temperature is higher. In this study no direct integration is included due to practical reasons in the

design of the heat recovery system.

Figure 2: Composite Curves and Grand Composite Curve of the theoretical potential of the system if solar

heating is not possible and the ambient air temperature is low (A, upper two curves), and the practical potential

for average ambient conditions, i.e. with some solar heating or when the ambient air temperature is high (B,

lower two curves).

3. Method

The drying system has been modelled in an in-house simulation model built in MS Excel, using CoolProp

thermodynamic database (Bell et al., 2014) for humid air properties together with mass and energy balance

calculations. Based on the modelling an assessment of the Total Lifetime Cost (TLC) for different heat recovery

loads have been made. This included a future scenario analysis for the electricity and biomass prices using the

Energy Price and Carbon Balance Scenarios tool ENPAC (Axelsson et al., 2018). The tool estimates the price

escalation from 2025-2050 for electricity and low-graded biomass for different scenarios for European

conditions. The scenario used in this paper is sustainable development (SD) as defined by the IEA in World

1389

Energy Outlook (IEA, 2018). The values in ENPAC were compared to the 2019 values of electricity and

biomass prices in Spain as stated by partners within the DRALOD project, i.e. 140 €/MWh-el and 9 €/MWh-

fuel.

Table 1: Electricity and biomass prices estimated using the tool ENPAC, with 2018 as reference year.

Year Unit

Electricity Price,

Sustainable

Development

Low-grade Wood Fuel

Price, Sustainable

Development

2018 €/MWh 44 14

2025 €/MWh 48 22

2030 €/MWh 48 31

2040 €/MWh 63 44

Meteonorm Timeseries (Meteotest, 2018) were used for the hourly climate data for an average year in Madrid.

The data was aggregated into 79 intervals since a lot of the hourly input data were similar. By comparing the

aggregated sample with the original 8,760 data points a deviation in output, as ton BSG fed to the dryer per

year, between the sample and the original data set was below 2 %.

3.1 Simulations of the dryer plant

The drying plant system without Heat Recovery and five systems with different sizes (0.5, 1, 2, 3, 4 MW) of heat

pumps were modelled and evaluated from a cost perspective by adding the yearly costs for the alternatives

during the economic lifetime of the plant (20 years). The aim of evaluating different sizes of heat pumps was to

make an estimation of the economic sustainability of installing an HRU. For each case and at each data point

the demand for each energy source, i.e. solar, Heat Recovery and biomass, was calculated. For all cases and

at each design point the full energy potential from the solar roof was used as primary heating source. The energy

delivered from the solar roof was calculated using the ambient temperature and relative humidity, the solar

irradiation and a volumetric air flow of 180,000 m3/h. The load delivered from the heat recovery system varied

dependent on the ambient conditions, amount of heat recovered, i.e. from 0 to 4 MW, and on the amount of

heat delivered by the solar roof. If the demand of the dryer was covered without using the full potential of the

heat pump the model assumes that the remaining heat is emitted. To cover the peak load the boiler was

designed to ensure that the entire demand was covered at each data point.

3.2 Techno-economic assessment

Total Lifetime Cost (TLC) of the different system designs were calculated as a summation of the Total Annual

Cost (TAC).

𝑇𝐿𝐶 = (∑ 𝑇𝐴𝐶 𝑦
𝑁

𝑦=1

) (1)

Where y denotes a year from 1 to N, for the economic lifetime of the plant. N = 20 years in this assessment.

The TAC was calculated as a summation of the hourly cost for each hour of operation during one year, Eq(2).

𝑇𝐴𝐶 = (∑ 𝑇𝐶𝑖
𝑀

𝑖=1

× 𝐼𝑛𝑑𝑒𝑥𝑖 ) × 𝐴𝑂𝑇/8,760 (2)

Where i denotes a datapoint from 1 to M, where M is the total amount of datapoints, i.e. 79, and Index represents

the size of datapoint i, i.e. the number of hours aggregated in the datapoint. AOT is the annual operating time

of the plant, which is 8,060 hours in this assessment. Since the hourly cost has been estimated for 8,760 time

points the AOT is divided by 8,760 to account for the availability of the plant on a yearly basis. This is done since

the exact time points where the plant is down are unknown. The Total Cost at datapoint i, 𝑇𝐶𝑖 , was defined as

in Eq(3). The size of the HRU and the boiler were set to cover the peak demand during the year.

𝑇𝐶𝑖 = (𝐼𝑆𝑐𝑟𝑏,𝑚𝑎𝑥
𝑖 + 𝐼𝐻𝑃,𝑚𝑎𝑥

𝑖 + 𝐼𝐵𝑜𝑖𝑙𝑒𝑟,𝑚𝑎𝑥
𝑖 ) × 𝐶𝑅𝐹/8,760 + 𝐸𝐶𝐻𝑃

𝑖 + 𝐹𝐶𝐵𝑜𝑖𝑙𝑒𝑟
𝑖 (3)

1390

Where 𝐸𝐶𝐻𝑃
𝑖 and 𝐹𝐶𝐵𝑜𝑖𝑙𝑒𝑟

𝑖 denotes the electricity cost for the heat pump and the fuel cost for the boiler at

datapoint i. 8,760 refers to the total number of hours per year in the calculation and the Capital Recovery Factor

(CRF) for the system investments was estimated according to Eq(4). Total investment cost has been estimated

per hour by dividing CRF with 8,760.

𝐶𝑅𝐹 =
𝑟(1 + 𝑟)𝑛

(1 + 𝑟)𝑛 − 1
(4)

Where r denotes the interest rate, 9 %, and n the economic lifetime, 20 y. The investment cost for the

prefabricated boiler was calculated using the almost linear correlation bellow

𝐼𝐵𝑜𝑖𝑙𝑒𝑟 (€) = 1,000 × (125 + 0.2357 × 𝑐𝑎𝑝𝑎𝑐𝑖𝑡𝑦) × 𝐿𝐹𝐵𝑜𝑖𝑙𝑒𝑟 (5)

For calculating the investment cost of the heat pump Eq(6) was used, and Eq(7) was used for the scrubber. The

Chemical Engineering Plant Cost Index (CEPCI) for 2007 (525.4) and 2018 (603.1) were used. The investment

cost of the heat pump was calculated as a sum of capital cost for the condenser, evaporator and compressor.

The investment costs for the condenser and evaporator were based on the largest heat exchanger area and the

investment cost for the compressor was based on the data point with the largest electricity requirement. For the

scrubber the investment cost was based on the volume flow of the wet exhaust gases.

𝐼𝐻𝑃(€) =
𝐶𝐸𝑃𝐶𝐼2018

532.9
× (26,000 + 2,700 × 𝑃𝑚𝑎𝑥

𝑖 0.75) ×
𝑈𝑆𝐷

𝐸𝑈𝑅
× 𝐿𝐹𝑐𝑜𝑚𝑝𝑟 +

100

0.09
× ( 𝐴𝐶𝑜𝑛𝑑,𝑚𝑎𝑥

𝑖 0.778 +

𝐴𝐸𝑣𝑎𝑝,𝑚𝑎𝑥
𝑖 0.778) ×

𝑈𝑆𝐷

𝐸𝑈𝑅
× 𝐿𝐹𝐻𝑋

(6)

𝐼𝑆𝑐𝑟𝑏 (€) = 37,000 ×
𝐶𝐸𝑃𝐶𝐼2018

𝐶𝐸𝑃𝐶𝐼2007
× (

𝑄𝑚𝑎𝑥
𝑖

𝑄𝑟𝑒𝑓
)

0.7

×
𝑈𝑆𝐷

𝐸𝑈𝑅
× 𝐿𝐹𝑆𝑐𝑟𝑏 (7)

Where 𝑄𝑟𝑒𝑓 = 36,000 m
3/h, USD/EUR = 8/10, 𝑃𝑚𝑎𝑥

𝑖 installed electricity power in kW and Ai in m2. The Lang factor

of the boiler (LFBoiler) is set to 2, the Lang factor of the compressor (LFCompressor) is set to 2.5, the Lang factor of

the scrubber (LFScrubber) is set to 3.5 and the Lang factor of the heat exchanger (LFHX) is set to 3.5.

4. Result

The results for variating the size of the heat pump for different energy markets scenarios are shown in Figure 3.

Figure 3: Comparison in TLC for ENPAC Scenario and Spain current prices for different HRU, 0, 0.5 1, 2, 3 and

4 MW. The solar roof is excluded from the TLC.

Figure 3 shows that the TLC of the system without heat recovery is higher for the ENPAC scenario than the

Spain current prices, and that the TLC decreases for the ENPAC scenario and

1391

increases for Spanish scenario with increased heat recovery capacity. For ENPAC the figure also indicates that

the decrease in TLC after 3 MW flattens out. The reason that the trend flattens after 3 MW depends on two

things. The first reason is that the model is based on the assumption that the air flow into the scrubber is fixed,

which gives that the COP decreases for larger systems since the temperature lift needs to be higher as more

heat is extracted from the scrubber. The second reason can be understood by comparing the temperature out

of the solar roof for the 79 aggregated data sets and the corresponding heat demand for each of these sets. For

some sets a small load is required to meet the demand, decreasing the time the heat pump can operate at full

potential, in this case from 100 % to 45 % between 0 and 4 MW installed capacity. These factors the result is

also dependent on the climate data and the relationship between fuel cost and electricity cost (i.e. CFuel/CElectricity).

For an HRU to be relevant in the simulated system the CFuel/CElectricity needs to be below 5. The estimated prices

for electricity and biomass in Spain, 9 €/MWh for biomass and 140 €/MWh for electricity, indicate that

implementation of a heat recovery system is not economically viable. Comparing the electricity prices used in

this study with the electricity price in 2019 for European counties from Eurostat it can be seen that there is a

large variation and uncertainty related to the estimation, and a more thorough assessment would be needed in

order to draw more specific conclusions on making investments in heat recovery (Eurostat Statistics Explained,

2019).

5. Conclusions

This paper has analysed the integration of a Heat Recovery Unit into a system comprising of both solar and

biomass-based preheating of air to a low temperature drying plant A model has been developed and used in

order to assess the economic potential of the HRU for a plant based in Madrid, Spain. The results indicate that

for Spanish conditions the HRU is not likely to be feasible compared to a biomass-based hot water boiler. The

study reveals that for integration of the Heat Recovery Unit to be interesting from an economic perspective a

low electricity cost is needed in relation to the biomass cost. The relationship between fuel cost and electricity

cost (i.e. CFuel/CElectricity) needs to be below 5 in the specific study conducted. This ratio will depend on the

location of the drying system due to the variations in meteorological conditions. The designed model can be

applied to other locations in order to study the potential benefits of heat recovery, but then the input data needs

to be changed (i.e. meterological information and economic inputs like electricity and biomass prices). Future

work to improve the accuracy of the model and assessment includes addition of detailed knowledge of the dryer

performance and an optimization of the operational strategy for the entire system throughout the year.

Acknowledgements

This research has been supported by the EU project “Fast Track to Innovation - FTI”, DRALOD is co-funded by

the European Commission - Horizon 2020 Programme - under Grant Agreement 820554. The DRALOD project

is a collaboration between PERNIA, DBFZ, ÖKOTHERM and RISE. For more information about DRALOD

please visit dralod.com.

References

Alves-Filho O., 2018, Energy effective and green drying technologies with industrial applications. Chemical

Engineering Transactions, 70, 145–150.

Axelsson E., Bisaillon M., Löfblad, E., 2018, ENPAC 2.1 – update of the wood fuel module, Profu.

Bell I.H., Wronski J., Quoilin S., Lemort V., 2014, Pure and pseudo-pure fluid thermophysical property evaluation

and the open-source thermophysical property library coolprop. Industrial & Engineering Chemistry

Research, 53(6), 2498–2508.

Eurostat Statistics Explained, 2019, Electricity price statistics <ec.europa.eu/eurostat/statistics-

explained/index.php/Electricity_price_statistics> accessed 29.04.2020.

Grubbauer A., Fluch J., Brunner C., Ramschak T., Wilk V., Fleckl T., 2018, Renewable and highly efficient

energy systems through innovative combination of solar thermal and heat pump systems. Chemical

Engineering Transactions, 70, 745–750.

IEA, World Energy Outlook, 2018, International Energy Agency <iea.org/reports/world-energy-outlook-2018>

accessed 29.04.2020.

Marechal F., Muller D., 2008, Energy management methods for the food industry, Handbook of Water and

Energy Management in Food Processing. Woodhead Publishing/Elsevier, 221–255.

Meteotest, 2018, Meteonorm. <meteonorm.com/en/> accessed 29.03.2020.

Schosser F., Arpagaus C., Wamsley T. G., 2019, Heat pump integration by Pinch Analysis for industrial

applications: a review. Chemical Engineering Transactions, 76, 7–12.

1392

https://www.oeko-therm.net/en/