Designing Solar with Storage Systems for Energy Independence with Own Power 24 x 7

With the rise of solar power, the typical Grid-Tied solar systems are increasingly graduating to Hybrid systems coupled with Storage for the following reasons:

• The tariff for grid supply is increasing with the introduction of dynamic tariffs based on Time of Day (TOD) in India from 1 April 2024 to incentivize the use of stored solar during peak hours.

• Energy independence with stored solar power avoids expensive peak hours tariffs.

• It avoids loss of productivity during power cuts

• Avoids additional CAPEX and huge OPEX for an alternate power backup (Diesel Genset).

• The price drop in Lithium-ion LiFePO4 (LFP) Battery Packs makes storage more affordable.

With the addition of Battery Packs coupled with Hybrid Inverters, the design of such systems needs a full understanding of the concepts and professional design practices. To meet the various design criteria, the Hybrid Inverters provide many features and options.

Safety and performance are the most important aspects of the battery packs. Most lead-acid battery based systems were small and simple, with just one or two batteries. However, as system sizes became larger, many batteries needed to be used. Many such installations do not have a Battery Management System (BMS), which led to many major safety issues and poor performance.

This article will discuss all design aspects of Solar systems with Hybrid Inverters using the popular LFP Battery Packs. The design philosophies discussed here apply equally to pure off-grid systems (stand-alone without grid) and pure DC micro-grid systems.

1. TOD and the need for Energy Storage System (ESS)

The Time Of Day (TOD) billing will be effective nationwide in India starting from April 1, 2024, per the Ministry of Power, Government of India’s directive for managing the demand side of the National Electricity Grid. To facilitate this, smart meters are being installed for all commercial and industrial (C&I) consumers. During peak hours (6 AM to 10 AM and 6 PM to 10 PM), the tariff for C&I consumers will increase by approximately 25% starting from April 1, 2024. However, during off-peak hours (10 PM to 6 AM), the tariff for C&I consumers may decrease by about 10%.

Many commercial entities conduct most of their business during evening peak hours, leading to high electricity costs. To deal with the unreliability of electricity, every business has a Diesel Generator (DG) backup. Our survey indicates that commercial entities use DG for at least one hour daily. The cost of diesel power and the maintenance of DG sets are additional burdens. Between 20 to 50% of consumption happens during Solar hours (6 AM to 6 PM).

The harvested solar energy can be directly used to power the loads during peak hours. The LCOE (Levelized Cost Of Energy) of Solar is only around INR 5 per unit for plants up to 5KW.

2. Significance & Relationship between Sizing & Lifecycle

The life of the battery pack depends on two important factors.

• Rate of Charging & Discharging – Lower the Rate, the higher the battery life

• Depth Of Discharge (DOD) – Lower the Depth, the higher the battery life

The Rate of Charging is controlled by design. The design needs to ensure adequate Solar capacity for at least 6 solar hours (say, from 9:00 AM to 3:00 PM) to support the load directly and charge the battery. With six hours of charging, the Rate of Charging is 1/6 or 0.1666 C.

The Rate of Discharge is determined by the total load duration to be supported by the Storage/Battery. The lower the duration, the lower the pack capacity and the higher the discharge current. If the storage duration is only two hours, discharging will happen at 1/2 or 0.5 C. Instead, if the storage duration is ten hours, discharging will happen at 1/10 or 0.1 C (assuming full and flat load).

With a 0.25C rate for both charging and discharging and 90% DoD restriction, an LFP Battery can support nearly 6500 Cycles. By following the design principles mentioned on the next page, up to 9000 cycles (25 years) can be supported by an LFP Pack.

• Restrict the DoD to 90% with the help of BMS.

• Due to the inherent nature of solar coupled systems (for stationary applications), limit the charging current rate (C-rate) to 0.166C or less from 9.00 AM to 3.00 PM.

• Design to support for at least six hours for the full load (or as long the duration as possible) for ensuring 0.166C or less rate.

Manufacturers themselves provide 10 years of warranty (3650 cycles) for LFP Packs designed (restricting DoD to 90%) and manufactured specifically for Stationary applications.

3. Energy System Components Selection & Sizing

The System Sizing completely depends on the total load and duration/hours of load to be powered by Solar + Storage. The current prices of LFP batteries are attractive, causing the LCOS (Levelized Cost of Storage) to be very comparable to the electricity costs by DISCOMs during peak hours.

Levelized Cost of Total Energy (LCOTE) = Weighted Average of LCOE + LCOS

LCOE (during Solar hours); LCOS (during non-Solar hours, powered by Stored Solar)

Case of a commercial entity consuming 40 units every day:

This entity consumes about 12 units during the day (6 AM to 6 PM) but at its peak, it consumes about 27 units between 6 PM to 10 PM. During the night from 10 PM until early the next morning at 6 AM, it only consumes about one unit.

During solar hours, the entire 12 units can be directly harvested from Solar, benefitting the maximum (LCOE approximately INR 5 per unit). For the remaining 28 units, Storage can be used to store the Solar Power harvested during Solar hours and used during non-Solar hours (LCOS approximately INR 12.59 per unit).

LCOE and LCOS are integrated in a sophisticated manner, resulting in lesser costs than the Electricity Charges charged by DISCOMs. The proposition becomes even more attractive when Diesel Power’s cost is accounted for, albeit the usage is as little as one hour per day.

3a. Efficiency of the Entire System

While accounting for the Efficiency of the Inverter, there are three components of losses to be accounted for:

a) Losses during Charging from Solar (via CC-MPPT) – During Charging, only six hours of Solar energy are accounted for (from 9 a.m. to 3 p.m.); hence, 85% of the Solar Energy is captured as per empirical data. The remaining 15% of solar energy is either supplied to the load (along with stored solar from the battery) or float charges the battery if the load is less (or compensates for the reduction in yield on rainy days). Hence, losses are negligible during charging.

b) Losses in the Battery (only about 1%) – The LFP packs will not easily lose the stored charge, offering 99% efficiency. This efficiency varies with chemistry; for Lead Acid, it is about 85% (and much less for Deep Cycles or higher C-rate).

c) DC to AC conversion by Inverter – DC to AC conversion efficiency is typically specified as per the Euro efficiency curves of the Inverter and can be generally taken as 95%.

The above losses can be taken together as 8% (for a small-sized system), offering 92% efficiency.

3b. Overall System Sizing and Design

The size/design of the solar system, battery pack, and inverter needs to be calculated. As the first step, the following data are to be extracted from the Electricity Bill of the entity:

• The Sanctioned Load in KW (SL)

• Number of Energy Units Consumed (A)

• Duration of the Billing (B)

• The Maximum Demand (MD) recorded

• Consumption per day, C = A/B

3c. Battery Sizing

Consumption during Solar hours (D) Size of the Storage required,

E = C-D i.e. 40-12 = 28 units

• As detailed earlier, the Efficiency (ήess) of the Energy Storage Systems (ESS) is about 92%.

• Energy required for Storage, Eess = E/(ήess), i.e. 28/0.92 = 30.43 units of Solar Energy.

• We need to harvest solar energy and store Eess (30.43) KWh of energy to obtain 28 units of energy for the load.

• Typically, the DOD needs to be typically restricted to 90%.

• The size of the Pack (Psize) = Eess/DOD, i.e. 30.43/0.9 = 33.82 KWhr

However, Battery Packs are manufactured only in certain sizes: typically 5.0 KWh, 10 KWh, 15 KWh, and so on. Therefore, for the Psize, we need to choose the Pstd-size.

In this case, for 33.82 KWhr, one can select between 30 KWh, 35 KWhr, or 40 KWh.

A fine balance needs to be made for this choice. Battery Packs are the most costly item; if we select higher capacities (35 or 40 KW), and the ROI will also be lesser. However, if we select a slightly lower capacity Pack (30 KWh, in this case), the full Battery capacity shall be almost always utilized, and the CAPEX will be lesser to that extent.

3c.1 Grid as Backup Supply

Being close to the equator, most parts of India enjoy 330 to 340 days of good solar insolation per year. The average yield is 4.2 units per kilowatt of solar generation per day. However, during rainy or cloudy days (around 30 to 40 days a year), the solar insolation is insufficient to generate the required solar energy. Therefore, on low-yield days, power needs are met from the grid supply. Another reason for having grid supply as a backup is to account for the decreased capacity of solar packs over the years until battery augmentation (explained later).

3c.2 Number of Days of Autonomy

The only way to avoid the Grid supply is to increase the number of days of Autonomy from the default one day to two days or more, which means the size of the Pack needs to be doubled or more. For example, instead of 30KWh, the Pack size becomes 60KWh or more, which is very, very expensive. Hence, except for Off-Grid systems, the number of days of Autonomy is only one day by default.

3c.3 Battery Sized with Std sizes and the Available Capacity of Stored Energy

With the definite use of Grid supply, there will be fixed* charges from DISCOMs for every KW of Sanctioned Load. Therefore, it is prudent always to consume a little from Grid. It is a good compromise to size the Pack to the nearest std-under-size, hence we choose 30 KWh size.

In this case, for the selected 30 KWh Pack with 90% DOD, Eess available will be 30 x 0.9 =27 KWh only. Further, the ήess will bring down the available Energy (Eavail) for the Load to (27 x 0.92) = 24.84 KWh.

*One possibility to minimize fixed charges is to limit the number of connections in the premises. For example, if there are four connections in a premises, only two are adequate, as each connection can support two systems.

3c.4 Battery Chemistry

Lead-acid batteries work best at 50% depth of discharge or less but degrade quickly, require frequent replacements, and take up a lot of space. On the other hand, LFP (LiFePO4) packs offer better performance, built-in safety features, and more affordable pricing. Sodium-ion batteries are also expected to emerge as a preferred technology for stationary storage systems in near future.

3c.5 Low Voltage (48V) vs Higher Voltage Packs

The choice of voltage for the battery pack is a crucial factor. Deciding whether the pack should be a low-voltage (LV) pack, typically 48 volts, or a high-voltage (HV) pack is important. The LV pack offers several advantages, including lower cost, as the industry has matured significantly for the mobility market and has optimized the production of LV packs. However, LV packs have a limit on the total energy they can store.

Beyond this limit, the busbars become heavy and cumbersome, increasing costs and decreasing efficiency at high power levels. Generally, LV packs can manage up to about 50 to 100 kWh. On the other hand, HV packs require inverters that can handle high battery voltages, which can be expensive. However, as the market develops, we can expect the prices of HV packs to decrease, similar to what has happened with solar and LV packs. It is inevitable to use HV packs as the demand for handling higher power and energy increases. Generally, the higher the power and energy, the higher the voltage required (up to 1400 volts for MWh scales).

3d. Solar Sizing

“The solar capacity needs to cater to two categories: 1) Direct to Load and 2) Stored & Used. In the same example case as above, the direct-to-load is 12 kWh. Based on the prescribed size, we can store 27 kWh (in a 30 kWh pack) to obtain 24.84 kWh of stored energy. The remaining energy (40 – 12 – 24.84 = 3.16 kWh) will be drawn from the grid. Therefore, solar needs to cater to a total of 12 + 27 = 39 kWh. At an average solar generation of 4.2 kWh per kW, the solar capacity required is 9.07 kWp.”

Similar to the Pstd-size decision, the Solar String design also needs to be adjusted with the available Solar Modules’ capacity and matching characteristics to form strings suitable to the Inverter’s MPPT Voltage range and the Voc, not exceeding the permissible Inverter’s Max voltage. It is recommended to design the Strings for slight over-sizing of the nearest Solar capacity to compensate for the yearly degradation of Solar.

3e. Inverter Sizing & Max Demand

Most reputable brands specify the output capacity in kilowatts (KW). If the output capacity is mentioned in kilovolt-amperes (KVA), it’s important to consider the KVA to KW conversion factor of approximately 0.7. This means a 10 KVA inverter can only power about 7.0 KW load. Additionally, inverter specifications typically include information about overloading capacity for different durations:

• 1.13 (113%) times the rated power for 30 minutes

• Two (200%) times the rated power for 10 seconds

• Three (300%) to Five (500%) times the rated power for 10 to 20 milliseconds, which helps handle in-rush currents.

It is a common practice to double the size of the inverter to accommodate two factors:

1. To handle unintended overloading and in-rush current

2. To meet a small percentage of additional consumption in the near future.

Following this thumb rule usually ensures that the Maximum Demand (MD) will be within the capabilities of the inverter ratings. However, double-checking that the MD is within the chosen inverter rating is always a good practice.

3e.1 Suitability with Battery Chemistry

The legacy Inverters worked only with Lead-Acid Batteries. Modern Inverters support Lead-Acid, LFP and a few other chemistries also.

3e.2 Hybrid Inverter can act as Grid-Tied Inverter

It is brilliantly possible to use Hybrid Inverters as Grid-Tied Inverters also, simply by treating the Hybrid Inverter technically like the Grid-Tied Inverter at both Input(s) and Output. Hence, adding Battery Packs later without replacing the Inverter is a great possibility that saves money and effort. Therefore, it is recommended to use a Hybrid Inverter instead of just a Grid-tied Inverter if Battery Packs are to be added in the future.

3e.3 3-Phase or 1-Phase

Beyond a certain size (say, 5KW), the loads will invariably be 3-phase. Up to 8.0 KW, Single-phase Inverters are available, and they are at lesser costs than 3-phase Inverters. At and above 8.0 KW, practically, the systems (and Loads) will be 3-phase only. The 3-phase Inverters can also be used for 1-phase loads by balancing the loads. Asymmetric loading between the phases of up to 30 to 35% is also harmoniously handled by most good quality inverters.

3e.4 Paralleling the AC Outputs of Inverters

You can enhance the total capacity by connecting the AC outputs of multiple inverters in parallel. This applies to both single-phase and three-phase inverters. You can connect up to six inverters in parallel, and some models support up to 16 inverters in parallel. Many brands offer single-phase inverters, and you can configure three of these inverters to supply three-phase power with a perfect 120-degree phase shift between their outputs. For a three-phase supply, you can connect three inverters in each phase, totaling nine inverters for all three phases, in parallel.

The Inverter’s outputs are paralleled when the nearest Inverter capacity is NOT available for the required load (e.g., for a 20KW load, two 10KW Inverters can be used) and to facilitate future capacity expansion.

Whenever multiple Inverter outputs are connected in parallel, proper and guided configuration and communication between Inverters are to be established. Needless to say, it is always good to connect the same brand of Inverters.

3e.5 HF vs LF Inverters / Galvanic Isolation

The frequency of switching transistors distinguishes between low and high-frequency inverters. If the Switching happens at Line (Low) frequency, they are LF Inverters. The advancements in technology have brought the HF Inverters, wherein the switch devices operate at very high speeds/frequencies (10s of KHz to MHz), resulting in very small reactive components. These are also called transformer-less inverters, as the transformers become very small at high frequencies of operations and offer the same galvanic isolation.

Please note: Medical instruments need double-protected Galvanic isolation; to ensure the same, the instruments themselves will have their built-in (yet, this needs to be doubly ensured to support them).

3e.6 Matching Charging Current between Inverter and Pack’s Allowed Charging current

Care should be taken to confirm that the inverter’s Maximum Battery Charging Current is higher than the Pack’s designed/expected Charging Current when supplied from Solar (Grid, too). This would ensure the fully discharged Pack would charge 100% SOC with Solar Power by EOD.

3e.7 Permissible Solar Capacity || DC-Overloading

The same practices of DC overloading for the Grid-Tied Inverter can also be applied to Hybrid Inverters. In fact, Hybrid Inverters need to handle a much larger amount of solar power (up to 5.5 times the output capacity) to run solely on solar power for 24 hours. However, most current inverter datasheets only permit up to 1.5 times the capacity. One possible solution would be to use external Charge Controllers to support the required solar capacity and charge more packs, connecting these packs in parallel at the battery DC input of the inverter. However, it’s even more important to configure the Charge Controllers properly with the inverters and all other packs, which requires careful settings guided by the inverter manufacturer.

3e.8 Power Factor and Zero-Export feature of Inverter

Reactive Power Compensation is a standard feature in all Inverters. It needs to be set according to the Load’s Power factor, similar to what is usually set for an On-grid Inverter. Some Inverters have an Inbuilt zero export function to avoid Power Export to the Grid.

3e.9 Communication with BMS & APP

BMS and Inverter communicate either using the CAN bus (as is the norm in mobility) or the MOD bus, which is preferred for Stationary Storage systems. The Battery Temperature, SOC and other such parameters are conveyed over the bus. Both use RS485 Physical layer mostly with RJ-45 connectors.

Offering an app for remote monitoring of the LIVE status and past data (generation, load, battery health, etc.) stored in the cloud is becoming a default facility for all inverters for remote monitoring over the Internet.

4. Battery Packs: Connections, Degradation & Augmentation

In mobility, 80% to 70% of battery state of health (SOH) is considered the end of life (EOL). However, there is no defined EOL for stationary storage systems. Even as the battery capacity degrades to 70%, a significant amount of life remains in it. Therefore, instead of replacing the entire battery for full capacity, it would be more appropriate to supplement the lost capacity with an additional pack.

However, many technicalities must be understood/accommodated right from the system design to installation to be future-ready for Battery augmentation later. The most important is the number of Packs, as detailed below.

4a. Design of Number of Battery Packs

The lost battery capacity due to degradation can be augmented with additional Pack, subject to the following conditions:

• The pack used for augmentation has the same chemistry (LFP, etc.), voltage, AH capacity, and charging/discharging characteristics as the original packs.

• The BMS settings (e.g. max current during charging & discharge) should be identical.

• It is better to design with multiple packs rather than a single pack, as it would be far simpler to augment them as the capacity degrades, as explained below.

Hence, if four packs are used, even as the Pack degrades to, say, 80%, one more identical Pack can be added. The additional cost for Battery Augmentation will also be much less (than replacing entire Packs).

• However, in cases where only one pack is used, one need NOT wait for 50% degradation to Augment. Depending on the load’s demands/nature, it can be done at any time. Never augment with dissimilar Packs and ensure the BMS of all packs is identically set. Although it would be of excess capacity, this is the technically appropriate Solution.

Another possible solution: If Multi Inverters are used with their respective packs, the older packs can be clubbed for one set of Inverters. New packs for Augmentation can be used with another set of Inverters.

• With more than one pack, one of the Packs will need to be configured as Master BMS and others as Slave Packs. All of them will have a daisy-chained Communication loop with CAN or MOD-Bus over RS-485, mostly with RJ-45 connectors. The purpose of this is to have a scalable architecture with a Modular design for the Packs to connect as many Packs as needed and also keep the wiring simple and easy to communicate the various Pack parameters.

• It is to be noted that all Packs will be identical, and any Pack can be configured as Master BMS. In some arrangements, the Physical location, say the bottom-most, of the Rack that accommodates Multiple Packs is designated as Master BMS.

5. Diesel Genset Usage

When the usage of diesel generators and their expenditures are accounted for, even for just one hour per day @INR 23 per unit of diesel power, the financials become very attractive, and the Project IRR improved by about 3%. When just one maintenance/shut-down day per month (causing six hours of Power cut) is additionally accounted for, the Project IRR improves by another 1%.

The DG Power cost can be higher, and hence, the financial benefits will also be higher. The O&M of the DG Sets is also very cumbersome, and switching to a Renewable Solution around the clock will free you from maintaining the costly and noisy DG set.

6. Concluding Remarks

It is a natural evolution to grow from On-Grid to Hybrid systems for harvesting more clean power 24×7, especially when it is affordable, and the clean Power costs are lesser than dirty Grid power. In a nutshell, by switching to Solar + Storage, for nearly 24 x 7 x 365, entities can save themselves from huge power costs and also avoid the need to buy and maintain DG sets for backup power.

Meet the authors

Ilangovan Angaiah – President, ESS division, Sodion Energy & AR4-Tech Pvt ltd, Coimbatore, TN. Sodion Energy’s Pack assembly line with LASER welding and automated production & Testing Line for the Sodium-ion & Lithium-ion Battery Packs. AR4-Tech is the marketing arm of Sodion Products.

Vigneswaran Velusamy – Founder, V-Mitra Energy Installers, Coimbatore, TN. Has been involved with many multi-MW ground-mounted systems (>100MW), rooftop systems, and solar water pumping systems (100+ installations) since 2015.